Machine learning
BUSINESS ENGINEERING – FINANCIAL
ENGINEERING
2024-2025
Data Science Challenge: 5
Exam: 15
- Multiple choice
➢ No guessing correction
➢ Censure: if 5 answers per question, you need to answer 60% correct to pass the multiple
choice part
➢ So answer all questions
➢ Details on the exam
- Open question
,Inhoudsopgave
LECTURE 1 – WHAT IS DATA SCIENCE AND MACHINE LEARNING? ....................................................................... 4
1 INTRODUCTION ............................................................................................................................................................................. 4
2 TERMINOLOGY.............................................................................................................................................................................. 4
2.1 Artificial Intelligence ........................................................................................................................................................... 5
2.2 Machine Learning ............................................................................................................................................................... 5
3 DATA SCIENTIST ........................................................................................................................................................................... 7
3.1 Importance ........................................................................................................................................................................... 7
3.2 Roles and skills .................................................................................................................................................................... 7
LECTURE 2 – MACHINE LEARNING .......................................................................................................................... 8
1 EXPLAINING VERSUS PREDICTING MODELLING ............................................................................................................................... 8
2 DATA PREPROCESSING ................................................................................................................................................................. 10
2.1 Sampling .............................................................................................................................................................................. 10
2.2 Encoding .............................................................................................................................................................................. 11
2.3 Missing values .................................................................................................................................................................... 12
2.4 Outliers ............................................................................................................................................................................... 12
2.5 Normalizing ........................................................................................................................................................................ 13
2.6 Discretization .................................................................................................................................................................... 14
3 SOME NOTES ABOUT CHATGPT ................................................................................................................................................... 14
LECTURE 3 – INTRODUCTION TO PREDICTIVE MODELING ............................................................................... 16
1. TERMINOLOGY ............................................................................................................................................................................ 16
2. FINDING INFORMATIVE VARIABLES FROM THE DATA ................................................................................................................... 17
3. DECISION TREES.......................................................................................................................................................................... 17
4. METHOLOGY OF DECISION TREES IN MORE DETAIL ..................................................................................................................... 18
5. OVERFITTING AND ITS AVOIDANCE ........................................................................................................................................... 20
5.1 Overfitting .......................................................................................................................................................................... 20
5.2 Avoidance ........................................................................................................................................................................... 21
5.3 Bias/ variance trade-off .....................................................................................................................................................22
LECTURE 4 – ASSESSING AND VISUALIZING MODEL PERFORMANCE ................................................................ 23
1. EVALUATING CLASSIFIERS ............................................................................................................................................................23
1.1 Accuracy .............................................................................................................................................................................. 23
1.2 Confusion matrix ................................................................................................................................................................ 23
1.3 Problems: unbalanced classes ............................................................................................................................................ 23
1.4 Problems: unequal costs and benefits .............................................................................................................................. 24
2. EXPECTED VALUE ....................................................................................................................................................................... 24
2.1 Expected Value for classifier evaluation .......................................................................................................................... 24
3. EVALUATION AND BASELINE PERFORMANCE .............................................................................................................................. 26
LECTURE 5 - LEVERAGING DATA SCIENCE: BUSINESS INSIGHTS, MODEL PERFORMANCE, AND EVIDENCE-
BASED DECISION MAKING .................................................................................................................................... 28
1 DECISION ANALYTIC THINKING I: WHAT IS A GOOD MODEL? (RECAP) ...................................................................................... 28
1.1 What is a Good Model? ...................................................................................................................................................... 28
1.2 Key concepts ...................................................................................................................................................................... 28
1.2.1 Positive and negatives ..................................................................................................................................................... 28
1.2.2 Accuracy .......................................................................................................................................................................... 28
1.2.3 The confusion matrix...................................................................................................................................................... 29
1.2.4 Cost- Benefit matrix ....................................................................................................................................................... 29
1.2.5 Expected Profit/ Value .................................................................................................................................................... 29
1.2.6 What is a good baseline?................................................................................................................................................ 29
2 VISUALIZING MODEL PERFORMANCE ........................................................................................................................................... 29
2.1 Ranking Classifier .............................................................................................................................................................. 30
2.2 Profit Curve........................................................................................................................................................................ 30
PAGINA 1
, 2.3 ROC Curve ......................................................................................................................................................................... 30
2.4 AUC (Area Under Curve)................................................................................................................................................... 31
2.5 Lift Curve (Cumulative Response Curve) ......................................................................................................................... 31
3 EVIDENCE AND PROBABILITIES.................................................................................................................................................... 34
3.1 Evidence .............................................................................................................................................................................. 34
3.2 Joint Probabilities and Independence............................................................................................................................... 34
3.3 Bayes Rule .......................................................................................................................................................................... 34
3.4 Naive Bayes and Conditional Independence ................................................................................................................... 34
3.5 Evidence Lift .......................................................................................................................................................................35
LECTURE 6 – SIMILARITY, NEIGHBORS, AND CLUSTERS .................................................................................... 37
1. SIMILARITY (BETWEEN INSTANCES) .............................................................................................................................................37
2. DISTANCE AND SIMILARITY MEASURES .......................................................................................................................................37
2.1 Euclidean distance ..............................................................................................................................................................37
2.2 Manhattan distance ...........................................................................................................................................................37
2.3 Cosine similarity ................................................................................................................................................................37
2.4 Jaccard similarity .............................................................................................................................................................. 38
2.5 Hamming distance ............................................................................................................................................................ 38
2.6 Levenshtein distance ........................................................................................................................................................ 38
3. K-NEAREST NEIGHBORS ............................................................................................................................................................. 38
4. HIERARCHICAL CLUSTERING (DENDROGRAMS) ......................................................................................................................... 40
5. K-MEANS CLUSTERING .............................................................................................................................................................. 40
LECTURE 7 – RECOMMENDER SYSTEMS ............................................................................................................... 41
1. WHAT IS A RECOMMENDER SYSTEM? ........................................................................................................................................... 41
2. PROBLEM DEFINITION: EVALUATING A RECOMMENDATION ALGORITHM .................................................................................. 41
3. RECOMMENDATION ALGORITHMS: TWO PERSPECTIVES ........................................................................................................... 42
3.1 Baselines, Content-Based, Collaborative Filtering and Hybrid Algorithms (Data perspective) .............................. 42
3.1.1 Baseline ............................................................................................................................................................................ 42
3.1.2 Content-based ................................................................................................................................................................. 42
3.1.3 Collaborative filtering ..................................................................................................................................................... 43
3.2 Pointwise, Pairwise and Listwise Learning-to-Rank (Learning perspective) ............................................................ 44
3.2.1 Pointwise: learning-to-rank ........................................................................................................................................... 44
3.2.2 Pairwise: learning-to-rank ............................................................................................................................................. 44
3.2.3 Listwise: learning-to-rank.............................................................................................................................................. 44
4. UNDER THE HOOD: BUILDING A PERSONALIZED RECOMMENDER SYSTEM ................................................................................ 44
LECTURE 8 – TEXT MINING .................................................................................................................................... 45
1. TEXT MINING APPLICATIONS ..................................................................................................................................................... 45
1.1 Unstructured vs. Structured Data ..................................................................................................................................... 45
1.2 Text Preprocessing............................................................................................................................................................. 45
1.3 Terminology: Documents, Tokens and Terms, Corpus ................................................................................................... 46
1.4 Bag of Words ..................................................................................................................................................................... 46
1.5 TF-IDF (Term Frequency - Inverse Document Frequency) ............................................................................................. 47
1.6 N-gram ............................................................................................................................................................................... 48
1.7 Named Entity Recognition ................................................................................................................................................ 48
1.8 Topic Model ....................................................................................................................................................................... 48
1.9 Word Embedding ............................................................................................................................................................... 49
2. ASSOCIATION RULE MINING...................................................................................................................................................... 49
2.1 Item sets ............................................................................................................................................................................. 49
2.2 Frequent Item sets............................................................................................................................................................. 49
2.3 Association Rules .............................................................................................................................................................. 50
2.4 Support .............................................................................................................................................................................. 50
2.5 Confidence ......................................................................................................................................................................... 50
2.6 Association Rule Mining: Apriori Algorithm .................................................................................................................. 50
2.7 Lift .......................................................................................................................................................................................52
PAGINA 2
,LECTURE 9 – NEURAL NETWORKS AND DEEP LEARNING ................................................................................... 53
1. NEURAL NETWORKS....................................................................................................................................................................53
1.1 The Perceptron ....................................................................................................................................................................53
1.2 Activation Function ............................................................................................................................................................53
1.3 Multi-Layer Perceptron ..................................................................................................................................................... 54
1.4 Forward Pass .......................................................................................................................................................................55
1.5 Loss Function ..................................................................................................................................................................... 56
1.6 Backpropagation (Backward Pass) .................................................................................................................................. 56
1.7 Gradient Descent Algorithm ............................................................................................................................................. 56
1.8 Stochastic Gradient Descent ............................................................................................................................................ 56
2. DEEP LEARNING......................................................................................................................................................................... 57
2.1 Convolutional Neural Networks (CNNs) ..........................................................................................................................57
2.2 Recurrent Neural Networks (RNNs) ................................................................................................................................ 58
2.3 Autoencoders ..................................................................................................................................................................... 58
2.4 Transformers ..................................................................................................................................................................... 59
2.5 Foundation Models (Large Language Models) ............................................................................................................... 59
LECTURE 10 – ENSEMBLE METHODS SVM ............................................................................................................. 61
1. ENSEMBLE METHODS ................................................................................................................................................................... 61
1.1 Combine by consensus ........................................................................................................................................................ 61
1.1.1 Bagging .............................................................................................................................................................................. 61
1.1.2 Random Forests ............................................................................................................................................................... 62
1.2 Combine by learning .......................................................................................................................................................... 62
1.2.1 Boosting ........................................................................................................................................................................... 62
1.2.2 Stacking ........................................................................................................................................................................... 63
2. A BRIEF INTRO TO THE SUPPORT VECTOR MACHINE ................................................................................................................. 64
3. DATA SCIENCE ETHICS ............................................................................................................................................................... 66
3.1 Data gathering: Privacy, A/B Testing and Bias ............................................................................................................... 67
3.1.1 Privacy .............................................................................................................................................................................. 67
3.1.2 Experimentation ............................................................................................................................................................. 67
3.1.3 Bias................................................................................................................................................................................... 68
3.2 Data preprocessing: Proxies, Government Backdoors ................................................................................................... 68
3.3 Modeling: ZK Proofs, Discrimination .............................................................................................................................. 69
3.4 Model evaluation: explain ................................................................................................................................................ 69
3.5 Deployment: Unintended consequences .......................................................................................................................... 70
PAGINA 3
,Lecture 1 – What is Data Science and Machine Learning?
What is machine learning? The automatic extraction of patterns from large amounts of data (done by machines).
Most of what people call AI = machine learning
1 INTRODUCTION
What will be the goals of a data-driven prediction? To find non- obvious patterns
= we use the patterns to improve our business (ex. To offer more product so they won’t be sold out)
Machine learning: automatic extraction of knowledge from data
Setting the scene with credit scoring example
o Banks: should I grand credit to this loan applicant?
o Predict the creditworthiness, based on historical data
Data → Machine learning technique → pattern
BUT note: an initial set of data instances with known target variable needed otherwise we can’t make a
predicted model!
Data instance xi
- A vector of size m (number of input variables)
- I = 1, 2, … ,m (number of data instances)
-
2 TERMINOLOGY
Machine learning: automatic extraction of patterns from data
Data science: a set of fundamental principles that guide the extraction of
knowledge from data
AI: methods for improving the knowledge or performance of an intelligent agent
over time, in response to the agent's experience in the world
Big data: data that is so large that traditional data processing systems are unable
to deal with it (both storage and analysis component).
= only a few companies have this (ex. twitter)
Querying and reporting (=displaying the data)
- You know exactly what you are looking for – targeted data extraction
- SQL – structured query language
Querying and Visualization (= graphical representation of data)
- Multidimensional analysis – exploring data across multiple dimensions (attributes and factors) to
uncover relationships, patterns and trends
PAGINA 4
, OLAP – Online analytical processing = a technology and methodology used to perform multidimensional analysis of
large datasets quickly and interactively.
! The end-user is still the engine of discovery
Business intelligence – getting the right information to the right person at the right time
It always includes these points :
Data is collected in a data warehouse. Tries to get all the data in
one place
Reporting uses data from the warehouse to generate visual
dashboards, tables, and summaries for business insights.
Machine learning uses the data in the warehouse as input for
training predictive models and uncovering patterns.
2.1 Artificial Intelligence
• A computer interacts through data
• Learning from data leads to intelligence
• Big Data + Machine Learning = Artificial Intelligence (Theodoros Evgeniou, 2019)
• Renewed interest from Deep Learning (large artificial neural networks)
• Most work in AI is on Machine Learning
• The automatic extraction of patterns from large amounts of data”
The separation between the fields has blurred
General definition: AI system leas a machine-based system that is designed to operate with varying levels of
autonomy and that may exhibit adaptiveness after deployment, and that, for explicit or implicit objectives, infers,
from the input it receives, how to generate outputs such as decisions, recommendations, predictions, classifications,
actions, or content creation (e.g., text, images, videos, or audio) to fulfill the intended purpose.
= The automatic extraction of patterns from large amounts of data
2.2 Machine Learning
Note: an initial set of data instances with known target variable
needed!!
Example Facebook likes predict personality traits
We can investigate the variables (likes) with highest and lowest
coefficient in the linear model.
Example. Default prediction with Facebook data for micro-finance
▪ In collaboration with NY-based LenddoEFL
▪ Mirco-finance in developing countries
▪ 1000 – 5000 $ loans
▪ Most citizens limited/no credit history
PAGINA 5
BUSINESS ENGINEERING – FINANCIAL
ENGINEERING
2024-2025
Data Science Challenge: 5
Exam: 15
- Multiple choice
➢ No guessing correction
➢ Censure: if 5 answers per question, you need to answer 60% correct to pass the multiple
choice part
➢ So answer all questions
➢ Details on the exam
- Open question
,Inhoudsopgave
LECTURE 1 – WHAT IS DATA SCIENCE AND MACHINE LEARNING? ....................................................................... 4
1 INTRODUCTION ............................................................................................................................................................................. 4
2 TERMINOLOGY.............................................................................................................................................................................. 4
2.1 Artificial Intelligence ........................................................................................................................................................... 5
2.2 Machine Learning ............................................................................................................................................................... 5
3 DATA SCIENTIST ........................................................................................................................................................................... 7
3.1 Importance ........................................................................................................................................................................... 7
3.2 Roles and skills .................................................................................................................................................................... 7
LECTURE 2 – MACHINE LEARNING .......................................................................................................................... 8
1 EXPLAINING VERSUS PREDICTING MODELLING ............................................................................................................................... 8
2 DATA PREPROCESSING ................................................................................................................................................................. 10
2.1 Sampling .............................................................................................................................................................................. 10
2.2 Encoding .............................................................................................................................................................................. 11
2.3 Missing values .................................................................................................................................................................... 12
2.4 Outliers ............................................................................................................................................................................... 12
2.5 Normalizing ........................................................................................................................................................................ 13
2.6 Discretization .................................................................................................................................................................... 14
3 SOME NOTES ABOUT CHATGPT ................................................................................................................................................... 14
LECTURE 3 – INTRODUCTION TO PREDICTIVE MODELING ............................................................................... 16
1. TERMINOLOGY ............................................................................................................................................................................ 16
2. FINDING INFORMATIVE VARIABLES FROM THE DATA ................................................................................................................... 17
3. DECISION TREES.......................................................................................................................................................................... 17
4. METHOLOGY OF DECISION TREES IN MORE DETAIL ..................................................................................................................... 18
5. OVERFITTING AND ITS AVOIDANCE ........................................................................................................................................... 20
5.1 Overfitting .......................................................................................................................................................................... 20
5.2 Avoidance ........................................................................................................................................................................... 21
5.3 Bias/ variance trade-off .....................................................................................................................................................22
LECTURE 4 – ASSESSING AND VISUALIZING MODEL PERFORMANCE ................................................................ 23
1. EVALUATING CLASSIFIERS ............................................................................................................................................................23
1.1 Accuracy .............................................................................................................................................................................. 23
1.2 Confusion matrix ................................................................................................................................................................ 23
1.3 Problems: unbalanced classes ............................................................................................................................................ 23
1.4 Problems: unequal costs and benefits .............................................................................................................................. 24
2. EXPECTED VALUE ....................................................................................................................................................................... 24
2.1 Expected Value for classifier evaluation .......................................................................................................................... 24
3. EVALUATION AND BASELINE PERFORMANCE .............................................................................................................................. 26
LECTURE 5 - LEVERAGING DATA SCIENCE: BUSINESS INSIGHTS, MODEL PERFORMANCE, AND EVIDENCE-
BASED DECISION MAKING .................................................................................................................................... 28
1 DECISION ANALYTIC THINKING I: WHAT IS A GOOD MODEL? (RECAP) ...................................................................................... 28
1.1 What is a Good Model? ...................................................................................................................................................... 28
1.2 Key concepts ...................................................................................................................................................................... 28
1.2.1 Positive and negatives ..................................................................................................................................................... 28
1.2.2 Accuracy .......................................................................................................................................................................... 28
1.2.3 The confusion matrix...................................................................................................................................................... 29
1.2.4 Cost- Benefit matrix ....................................................................................................................................................... 29
1.2.5 Expected Profit/ Value .................................................................................................................................................... 29
1.2.6 What is a good baseline?................................................................................................................................................ 29
2 VISUALIZING MODEL PERFORMANCE ........................................................................................................................................... 29
2.1 Ranking Classifier .............................................................................................................................................................. 30
2.2 Profit Curve........................................................................................................................................................................ 30
PAGINA 1
, 2.3 ROC Curve ......................................................................................................................................................................... 30
2.4 AUC (Area Under Curve)................................................................................................................................................... 31
2.5 Lift Curve (Cumulative Response Curve) ......................................................................................................................... 31
3 EVIDENCE AND PROBABILITIES.................................................................................................................................................... 34
3.1 Evidence .............................................................................................................................................................................. 34
3.2 Joint Probabilities and Independence............................................................................................................................... 34
3.3 Bayes Rule .......................................................................................................................................................................... 34
3.4 Naive Bayes and Conditional Independence ................................................................................................................... 34
3.5 Evidence Lift .......................................................................................................................................................................35
LECTURE 6 – SIMILARITY, NEIGHBORS, AND CLUSTERS .................................................................................... 37
1. SIMILARITY (BETWEEN INSTANCES) .............................................................................................................................................37
2. DISTANCE AND SIMILARITY MEASURES .......................................................................................................................................37
2.1 Euclidean distance ..............................................................................................................................................................37
2.2 Manhattan distance ...........................................................................................................................................................37
2.3 Cosine similarity ................................................................................................................................................................37
2.4 Jaccard similarity .............................................................................................................................................................. 38
2.5 Hamming distance ............................................................................................................................................................ 38
2.6 Levenshtein distance ........................................................................................................................................................ 38
3. K-NEAREST NEIGHBORS ............................................................................................................................................................. 38
4. HIERARCHICAL CLUSTERING (DENDROGRAMS) ......................................................................................................................... 40
5. K-MEANS CLUSTERING .............................................................................................................................................................. 40
LECTURE 7 – RECOMMENDER SYSTEMS ............................................................................................................... 41
1. WHAT IS A RECOMMENDER SYSTEM? ........................................................................................................................................... 41
2. PROBLEM DEFINITION: EVALUATING A RECOMMENDATION ALGORITHM .................................................................................. 41
3. RECOMMENDATION ALGORITHMS: TWO PERSPECTIVES ........................................................................................................... 42
3.1 Baselines, Content-Based, Collaborative Filtering and Hybrid Algorithms (Data perspective) .............................. 42
3.1.1 Baseline ............................................................................................................................................................................ 42
3.1.2 Content-based ................................................................................................................................................................. 42
3.1.3 Collaborative filtering ..................................................................................................................................................... 43
3.2 Pointwise, Pairwise and Listwise Learning-to-Rank (Learning perspective) ............................................................ 44
3.2.1 Pointwise: learning-to-rank ........................................................................................................................................... 44
3.2.2 Pairwise: learning-to-rank ............................................................................................................................................. 44
3.2.3 Listwise: learning-to-rank.............................................................................................................................................. 44
4. UNDER THE HOOD: BUILDING A PERSONALIZED RECOMMENDER SYSTEM ................................................................................ 44
LECTURE 8 – TEXT MINING .................................................................................................................................... 45
1. TEXT MINING APPLICATIONS ..................................................................................................................................................... 45
1.1 Unstructured vs. Structured Data ..................................................................................................................................... 45
1.2 Text Preprocessing............................................................................................................................................................. 45
1.3 Terminology: Documents, Tokens and Terms, Corpus ................................................................................................... 46
1.4 Bag of Words ..................................................................................................................................................................... 46
1.5 TF-IDF (Term Frequency - Inverse Document Frequency) ............................................................................................. 47
1.6 N-gram ............................................................................................................................................................................... 48
1.7 Named Entity Recognition ................................................................................................................................................ 48
1.8 Topic Model ....................................................................................................................................................................... 48
1.9 Word Embedding ............................................................................................................................................................... 49
2. ASSOCIATION RULE MINING...................................................................................................................................................... 49
2.1 Item sets ............................................................................................................................................................................. 49
2.2 Frequent Item sets............................................................................................................................................................. 49
2.3 Association Rules .............................................................................................................................................................. 50
2.4 Support .............................................................................................................................................................................. 50
2.5 Confidence ......................................................................................................................................................................... 50
2.6 Association Rule Mining: Apriori Algorithm .................................................................................................................. 50
2.7 Lift .......................................................................................................................................................................................52
PAGINA 2
,LECTURE 9 – NEURAL NETWORKS AND DEEP LEARNING ................................................................................... 53
1. NEURAL NETWORKS....................................................................................................................................................................53
1.1 The Perceptron ....................................................................................................................................................................53
1.2 Activation Function ............................................................................................................................................................53
1.3 Multi-Layer Perceptron ..................................................................................................................................................... 54
1.4 Forward Pass .......................................................................................................................................................................55
1.5 Loss Function ..................................................................................................................................................................... 56
1.6 Backpropagation (Backward Pass) .................................................................................................................................. 56
1.7 Gradient Descent Algorithm ............................................................................................................................................. 56
1.8 Stochastic Gradient Descent ............................................................................................................................................ 56
2. DEEP LEARNING......................................................................................................................................................................... 57
2.1 Convolutional Neural Networks (CNNs) ..........................................................................................................................57
2.2 Recurrent Neural Networks (RNNs) ................................................................................................................................ 58
2.3 Autoencoders ..................................................................................................................................................................... 58
2.4 Transformers ..................................................................................................................................................................... 59
2.5 Foundation Models (Large Language Models) ............................................................................................................... 59
LECTURE 10 – ENSEMBLE METHODS SVM ............................................................................................................. 61
1. ENSEMBLE METHODS ................................................................................................................................................................... 61
1.1 Combine by consensus ........................................................................................................................................................ 61
1.1.1 Bagging .............................................................................................................................................................................. 61
1.1.2 Random Forests ............................................................................................................................................................... 62
1.2 Combine by learning .......................................................................................................................................................... 62
1.2.1 Boosting ........................................................................................................................................................................... 62
1.2.2 Stacking ........................................................................................................................................................................... 63
2. A BRIEF INTRO TO THE SUPPORT VECTOR MACHINE ................................................................................................................. 64
3. DATA SCIENCE ETHICS ............................................................................................................................................................... 66
3.1 Data gathering: Privacy, A/B Testing and Bias ............................................................................................................... 67
3.1.1 Privacy .............................................................................................................................................................................. 67
3.1.2 Experimentation ............................................................................................................................................................. 67
3.1.3 Bias................................................................................................................................................................................... 68
3.2 Data preprocessing: Proxies, Government Backdoors ................................................................................................... 68
3.3 Modeling: ZK Proofs, Discrimination .............................................................................................................................. 69
3.4 Model evaluation: explain ................................................................................................................................................ 69
3.5 Deployment: Unintended consequences .......................................................................................................................... 70
PAGINA 3
,Lecture 1 – What is Data Science and Machine Learning?
What is machine learning? The automatic extraction of patterns from large amounts of data (done by machines).
Most of what people call AI = machine learning
1 INTRODUCTION
What will be the goals of a data-driven prediction? To find non- obvious patterns
= we use the patterns to improve our business (ex. To offer more product so they won’t be sold out)
Machine learning: automatic extraction of knowledge from data
Setting the scene with credit scoring example
o Banks: should I grand credit to this loan applicant?
o Predict the creditworthiness, based on historical data
Data → Machine learning technique → pattern
BUT note: an initial set of data instances with known target variable needed otherwise we can’t make a
predicted model!
Data instance xi
- A vector of size m (number of input variables)
- I = 1, 2, … ,m (number of data instances)
-
2 TERMINOLOGY
Machine learning: automatic extraction of patterns from data
Data science: a set of fundamental principles that guide the extraction of
knowledge from data
AI: methods for improving the knowledge or performance of an intelligent agent
over time, in response to the agent's experience in the world
Big data: data that is so large that traditional data processing systems are unable
to deal with it (both storage and analysis component).
= only a few companies have this (ex. twitter)
Querying and reporting (=displaying the data)
- You know exactly what you are looking for – targeted data extraction
- SQL – structured query language
Querying and Visualization (= graphical representation of data)
- Multidimensional analysis – exploring data across multiple dimensions (attributes and factors) to
uncover relationships, patterns and trends
PAGINA 4
, OLAP – Online analytical processing = a technology and methodology used to perform multidimensional analysis of
large datasets quickly and interactively.
! The end-user is still the engine of discovery
Business intelligence – getting the right information to the right person at the right time
It always includes these points :
Data is collected in a data warehouse. Tries to get all the data in
one place
Reporting uses data from the warehouse to generate visual
dashboards, tables, and summaries for business insights.
Machine learning uses the data in the warehouse as input for
training predictive models and uncovering patterns.
2.1 Artificial Intelligence
• A computer interacts through data
• Learning from data leads to intelligence
• Big Data + Machine Learning = Artificial Intelligence (Theodoros Evgeniou, 2019)
• Renewed interest from Deep Learning (large artificial neural networks)
• Most work in AI is on Machine Learning
• The automatic extraction of patterns from large amounts of data”
The separation between the fields has blurred
General definition: AI system leas a machine-based system that is designed to operate with varying levels of
autonomy and that may exhibit adaptiveness after deployment, and that, for explicit or implicit objectives, infers,
from the input it receives, how to generate outputs such as decisions, recommendations, predictions, classifications,
actions, or content creation (e.g., text, images, videos, or audio) to fulfill the intended purpose.
= The automatic extraction of patterns from large amounts of data
2.2 Machine Learning
Note: an initial set of data instances with known target variable
needed!!
Example Facebook likes predict personality traits
We can investigate the variables (likes) with highest and lowest
coefficient in the linear model.
Example. Default prediction with Facebook data for micro-finance
▪ In collaboration with NY-based LenddoEFL
▪ Mirco-finance in developing countries
▪ 1000 – 5000 $ loans
▪ Most citizens limited/no credit history
PAGINA 5
