À propos de ce cours

15,559 consultations récentes
Certificat partageable
Obtenez un Certificat lorsque vous terminez
100 % en ligne
Commencez dès maintenant et apprenez aux horaires qui vous conviennent.
Dates limites flexibles
Réinitialisez les dates limites selon votre disponibilité.
Niveau débutant

Accessible to business-side learners yet also vital to techies. Engage in the commercial use of ML – whether you're an enterprise leader or a quant.

Approx. 17 heures pour terminer
Anglais

Ce que vous allez apprendre

  • Participate in the application of machine learning, helping select between and evaluate technical approaches

  • Interpret a predictive model for a manager or executive, explaining how it works and how well it predicts

  • Circumvent the most common technical pitfalls of machine learning

  • Screen a predictive model for bias against protected classes – aka AI ethics

Compétences que vous acquerrez

Data ScienceArtificial Intelligence (AI)Machine LearningPredictive AnalyticsMachine Learning (ML) Algorithms
Certificat partageable
Obtenez un Certificat lorsque vous terminez
100 % en ligne
Commencez dès maintenant et apprenez aux horaires qui vous conviennent.
Dates limites flexibles
Réinitialisez les dates limites selon votre disponibilité.
Niveau débutant

Accessible to business-side learners yet also vital to techies. Engage in the commercial use of ML – whether you're an enterprise leader or a quant.

Approx. 17 heures pour terminer
Anglais

Offert par

Placeholder

SAS

Programme du cours : ce que vous apprendrez dans ce cours

Semaine
1

Semaine 1

5 heures pour terminer

MODULE 1 - The Foundational Underpinnings of Machine Learning

5 heures pour terminer
10 vidéos (Total 83 min), 6 lectures, 12 quiz
10 vidéos
P-hacking: a treacherous pitfall8 min
P-hacking: your predictive insights may be bogus8 min
P-hacking: how to ensure sound discoveries9 min
Avoiding overfitting: the train/test split8 min
Why ice cream is linked to shark attacks7 min
Causation is just a hobby -- prediction is your job6 min
The art of induction: why generalizing from data is hard6 min
Learning from mistakes: why negative cases matter5 min
Intro to the hands-on assessment (Excel or Google Sheets)11 min
6 lectures
Why this course isn't hands-on & why it's essential for techies anyway19 min
The Machine Learning Glossary (optional)10 min
One-question survey1 min
Complementary materials on p-hacking (optional)10 min
Correlation does not imply causation (optional)10 min
Data access for auditors (optional)10 min
11 exercices pour s'entraîner
Course overview: Machine Learning Under the Hood4 min
P-hacking: a treacherous pitfall2 min
P-hacking: your predictive insights may be bogus2 min
P-hacking: how to ensure sound discoveries4 min
Avoiding overfitting: the train/test split4 min
Why ice cream is linked to shark attacks4 min
Causation is just a hobby -- prediction is your job4 min
The art of induction: why generalizing from data is hard4 min
Learning from mistakes: why negative cases matter2 min
Intro to the hands-on assessment (Excel or Google Sheets)2 min
Module 1 Review30 min
Semaine
2

Semaine 2

3 heures pour terminer

MODULE 2 - Standard, Go-To Machine Learning Methods

3 heures pour terminer
12 vidéos (Total 107 min), 1 lecture, 11 quiz
12 vidéos
Business rules rock and decision trees rule13 min
Pruning decision trees to avoid overfitting12 min
DEMO - Comparing decision tree models (optional)13 min
Drawing the gains curve for a decision tree6 min
Drawing the profit curve for a decision tree6 min
Naïve Bayes11 min
Linear models and perceptrons6 min
Linear part II: a perceptron in two dimensions8 min
Why probabilities drive better decisions than yes/no outputs7 min
Logistic regression6 min
DEMO - Training a logistic regression model (optional)4 min
1 lecture
A powerful, helpful visualization of how decision trees work (optional)10 min
11 exercices pour s'entraîner
A refresher on decision trees2 min
Business rules rock and decision trees rule4 min
Pruning decision trees to avoid overfitting2 min
Drawing the gains curve for a decision tree2 min
Drawing the profit curve for a decision tree2 min
Naïve Bayes2 min
Linear models and perceptrons2 min
Linear part II: a perceptron in two dimensions4 min
Why probabilities drive better decisions than yes/no outputs4 min
Logistic regression4 min
Module 2 Review30 min
Semaine
3

Semaine 3

4 heures pour terminer

MODULE 3 - Advanced Methods, Comparing Methods, & Modeling Software

4 heures pour terminer
16 vidéos (Total 154 min), 2 lectures, 14 quiz
16 vidéos
Neural nets: decision boundaries & a comparison to logistic regression8 min
DEMO - Training a neural network model (optional)2 min
Deep learning9 min
Ensemble models and the Netflix Prize8 min
Supercharging prediction: ensembles & the generalization paradox12 min
DEMO - Training an ensemble model (optional)3 min
DEMO - Autotuning a machine learning model (optional)3 min
Compare and contrast: summary of ML methods8 min
Machine learning software: dos and don'ts for choosing a tool11 min
Machine learning software: how tools vary and how to choose one11 min
Model deployment: out of the software tool and into the field9 min
Uplift modeling I: optimize for influence and persuade by the numbers12 min
Uplift modeling II: modeling over treatment and control groups12 min
Uplift modeling III: how it works – for banks and for Obama15 min
Uplift modeling IV: improving churn modeling, plus other applications13 min
2 lectures
The generalization paradox of ensembles (optional) 10 min
Complementary readings on uplift modeling (optional) 10 min
14 exercices pour s'entraîner
How neural networks work5 min
Neural nets: decision boundaries & a comparison to logistic regression2 min
Deep learning2 min
Ensemble models and the Netflix Prize2 min
Supercharging prediction: ensembles & the generalization paradox4 min
Compare and contrast: summary of ML methods4 min
Machine learning software: dos and don'ts for choosing a tool2 min
Machine learning software: how tools vary and how to choose one2 min
Model deployment: out of the software tool and into the field2 min
Uplift modeling I: optimize for influence and persuade by the numbers2 min
Uplift modeling II: modeling over treatment and control groups2 min
Uplift modeling III: how it works – for banks and for Obama4 min
Uplift modeling IV: improving churn modeling, plus other applications4 min
Module 3 Review30 min
Semaine
4

Semaine 4

4 heures pour terminer

MODULE 4 – Pitfalls, Bias, and Conclusions

4 heures pour terminer
7 vidéos (Total 76 min), 8 lectures, 8 quiz
7 vidéos
Machine bias II: visualizing why models are inequitable8 min
Machine bias III: justice can't be colorblind13 min
Explainable ML, model transparency, and the right to explanation15 min
Conclusions on ML ethics: establishing standards as a form of social activism8 min
Pitfalls: the seven deadly sins of machine learning11 min
Conclusions and what's next – continuing your learning10 min
8 lectures
The original ProPublica article on machine bias10 min
Interactive MIT Technology Review article on disparate false positive rates10 min
Another interactive demo of machine bias (optional)10 min
Complementary reading on machine bias (optional)10 min
More on explainable ML and model transparency (optional)10 min
Tallying the positive and negative impacts of AI (optional)10 min
John Elder's top ten data science mistakes (optional)10 min
Further resources and readings to continue your learning (optional)10 min
8 exercices pour s'entraîner
Machine bias I: the conundrum of inequitable models4 min
Machine bias II: visualizing why models are inequitable2 min
Machine bias III: justice can't be colorblind4 min
Explainable ML, model transparency, and the right to explanation4 min
Conclusions on ML ethics: establishing standards as a form of social activism2 min
Pitfalls: the seven deadly sins of machine learning4 min
Conclusions and what's next - continuing your learning2 min
Module 4 Review30 min

Avis

Meilleurs avis pour MACHINE LEARNING UNDER THE HOOD: THE TECHNICAL TIPS, TRICKS, AND PITFALLS

Voir tous les avis

À propos du Spécialisation Machine Learning for Everyone with Eric Siegel

Machine Learning for Everyone with Eric Siegel

Foire Aux Questions

D'autres questions ? Visitez le Centre d'Aide pour les Etudiants.