Back To Course Home

Mastering Machine Learning Theory and Practice

0% completed

Getting started

Introduction to the Course

Introduction

The Basic Idea of Machine Learning History of Machine Learning Mathematical Formulation of the Basic Learning Problem Nonlinear Regression in High-Dimensions Recent Advances No Free Lunch, but Worth the Bite Quiz

Scientific Programming with Python

Programming Environment Basic Language Elements Code Efficiency and Vectorization Data Handling Image Processing and Convolutional Filters Quiz

Machine Learning with Sklearn

Overview Classification Using Multiple Machine Learning Models Performance Measures and Evaluations Data Handling Dimensionality Reduction, Feature Selection, and t-SNE Decision Trees and Random Forests Linear Support Vector Machines (SVMs)Nonlinear Support Vector Machines (SVMs)Validation of Support Vector Machines (SVMs)Challenge: Which Classifier Works Well?Solution: Which Classifier Works Well?Quiz

Neural Networks and Keras

Neurons and the Threshold Perceptron

Multilayer Perceptron (MLP) and Keras

Applying MLP to the MNIST dataset

Representational Learning

Introduction to Convolutional Neural Networks (CNNs)

Applications of Convolutional Neural Networks (CNNs)

Object Detection

Challenges in Machine Learning

Challenge: Choose the Best Model

Solution: Choose the Best Model

Regression and Optimization

Linear Regression and Gradient Descent Error Surface and Challenges for Gradient Descent Advanced Gradient Optimization (Learning)Regularization: Ridge Regression and LASSO Nonlinear Regression Applying Gradient Descent to Neural Network Models Stochastic Gradient Descent (SGD)Artificial Neural Networks Implementing a Multilayer Perceptron for the XOR Problem Automatic Differentiation Quiz

Basic Probability Theory

Random Numbers and Their Probability (Density) Function Moments: Mean and Variance Examples of Probability (Density) Functions Cumulative Probability Function and the Central Limit Theorem Density Functions of Multiple Random Variables How to Combine Prior Knowledge With New Evidence Markov Chain Monte Carlo Quiz

Probabilistic Regression and Bayes Nets

Probabilistic Models Learning in Probabilistic Models: Maximum Likelihood Estimate Probabilistic Classification Maximum A Posteriori (MAP) and Regularization with Priors Bayes' Nets: Multivariate Causal Modeling Probabilistic and Stochastic Neural Networks Quiz

Generative Models

Modeling Classes Supervised Generative Models Naive Bayes Self-Supervised Generative Models Generative Neural Networks Challenge: Sentiment Analysis Using Naive Bayes Solution: Sentiment Analysis Using Naive Bayes Challenge: K-Means Clustering Solution: K-Means Clustering Quiz

Cyclic Models and Recurrent Neural Networks

Overview Sequence Processing Basic Sequence Processing with MLP and RNN Gated RNN and NLP Other Gated Architectures and Attention Markov Random Field Model and the Hopfield Model The Boltzmann Machine The Restricted Boltzmann Machine Quiz

Reinforcement Learning

Formalization of the T-Maze Problem Representation of the T-Maze roblem Model-Based Reinforcement Learning Model-Free Reinforcement Learning Deep Reinforcement Learning Actors and Actor-Critics Reinforcement Learning in the Brain Quiz

Artificial intelligence, the Brain, and Our Society

Overview Different Levels of Modeling and the Brain Machine Learning and Artificial Intelligence The Impact of Machine Learning Technology on Society

Conclusion

Wrap Up Copyright Information

Solution: Choose the Best Model

Let's find out which model is the best.

We'll cover the following

Challenge
- Difference between the models
Solution
Explanation

Challenge

You are given three models named model1, model2, and model3. Your task is to identify the following:

The model that is a bestfit (best performing model)
The model that is an overfit
The model that is an underfit

Difference between the models

The three models were saved at different epochs ranging from $1$ to $200$ . We did this because we wanted to fetch three models that were the bestfit, underfit, and overfit models.

Get hands-on with 1400+ tech skills courses.