Introduction to Machine Learning with Python

Name: Introduction to Machine Learning with Python
Rating: 4.3 (81 reviews)

Unlocking the Power of Data: A Comprehensive Guide to Machine Learning with Python

Created byTensor Teach

Last updated 5/2024

English

English [Auto],

What you'll learn

Learn how gradient descent optimizes the parameters of a machine learning model
Fit a linear regression model to a dataset to predict the win rate of a MLB baseball team
Train a decision tree and random forest model to predict the credit risk of a borrower
Learn how k-means clustering works and apply it to a dataset of news headlines

Course content

4 sections • 24 lectures • 1h 59m total length

What Is Machine Learning?8:44
Explore the intuition of machine learning, including supervised, unsupervised, and semi-supervised learning, with examples like car sales predictions, clustering, and time series forecasting.
Line of Best Fit4:43
Illustrate how to fit a line of best fit with simple linear regression on synthetic data, explain residuals, and introduce mean squared error as the loss function.
Cost Function8:22
Learn how mean squared error defines the cost function for linear regression, distinguishing single-example loss from dataset-wide cost, and how gradient descent discovers the best m and b.
Gradient Descent12:12
Explore how gradient descent iteratively updates parameters using derivatives and a learning rate to minimize mean squared error, navigating convex bowls and nonconvex landscapes to approach the global optima.

What Is Regression?5:23
Explain regression as a supervised learning approach, from simple to multiple regression, and show how to fit models to predict continuous values using metrics like mean squared error and R-squared.
Exploring Our Dataset9:54
Analyze the Moneyball dataset to identify predictors of winning percentage by comparing on-base and slugging differences, perform train-test splits, and prepare simple and multiple linear regression models.
Fitting a Simple Linear Regression Model4:18
Fit a simple linear regression using on base percentage difference to predict win rate, assess correlation with slugging percentage difference, and compare training and test set performance.
Fitting a Multiple Regression Model5:12
Add a second feature, slugging percentage, to a regression model and examine correlations and multicollinearity to improve win-rate prediction with adjusted R squared.
Notebook & Dataset for Regression Section0:01

What Is Classification?3:35
Use supervised learning to assign data to categories, focusing on binary classification. Introduce cross entropy loss and common models like logistic regression, decision trees, and random forest.
Introducing Logistic Regression8:04
Discover how logistic regression uses a sigmoid to convert linear predictions into probabilities for binary classification, and how cross-entropy loss evaluates model error.
Classification Dataset Overview3:43
Explore the german credit dataset for a logistic regression classifier, examining features like checking status, credit history, loan purpose, amount, and employment, and transform them for binary classification.
Fitting Our Logistic Regression Model6:19
Transform categorical features into numeric values using one-hot encoding and create a 75/25 train-test split with scikit-learn. Train a logistic regression model and report training and test accuracy.
Evaluating Our Logistic Regression Model11:44
Evaluate a logistic regression model using precision, recall, and F1, detailing false positives and negatives, and show how balancing data improves recall for bad-credit-risk cases at the expense of accuracy.
German Credit Dataset0:01
Logistic Regression Notebook0:02
What Are Decision Trees?7:12
Explore how decision trees classify data from root to leaf using Gini impurity, train on credit data, and limit depth to reduce overfitting while preparing for random forests.
Training a Decision Tree Classifier5:34
Train a decision tree classifier by selecting the best feature with the Gini coefficient and split data to a max depth, then evaluate on the German credit data.
Random Forest6:42
Explore how a random forest combines multiple decision trees trained on bootstrap samples and random feature subsets to improve predictions through majority voting.
Decision Tree & Random Forest Notebook0:04

Read Before Beginning Section0:15
Clustering Notebook0:03
K-means Clustering Explained3:10
Demonstrate how K-means clustering groups data by features in an unsupervised setting, assigns points to centroids, and updates centroids as the mean of assigned points through iterations.
Fitting Our K-means Clustering Model4:23
Implement a k-means clustering workflow on OpenAI headline embeddings, reduce 1400+ dimensional vectors with PCA, and visualize five topic clusters in two-dimensional or three-dimensional space.

Requirements

Intermediate Python programming
Introductory understanding of Calculus, mainly derivatives
Introductory understanding of Linear Algebra, mainly what vectors are

Description

This course is designed to provide a thorough introduction to the world of machine learning. This course is perfect for beginners and those looking to enhance their data science skills using Python.

Section 1: Introduction to Machine Learning In this section, we will explore the fundamentals of machine learning. We'll start by defining machine learning and understanding its significance in today's data-driven world. We'll walk through a simple example, such as finding the line of best fit, to illustrate core concepts. Key topics like cost functions and the optimization technique of gradient descent will be covered, along with understanding the importance of the learning rate.

Section 2: Regression Regression analysis is a powerful tool for predicting continuous outcomes. We'll dive into different regression models and learn how to evaluate their performance. You'll gain hands-on experience by exploring datasets and fitting both linear and multiple regression models. This section ensures a solid foundation for understanding how regression works and how to apply it effectively.

Section 3: Classification Classification techniques are essential for predicting categorical outcomes. We'll begin by explaining what classification is and introducing logistic regression. You'll work with datasets to fit logistic regression models and understand their applications. The section also covers advanced techniques like decision trees and random forests, providing a comprehensive understanding of various classification methods.

Section 4: Clustering Clustering helps in grouping data points with similar characteristics. We'll focus on the K-means clustering algorithm, starting with an overview of the method. You'll learn how to explore datasets and fit clustering models to uncover hidden patterns and insights within your data.

By the end of this course, you'll be equipped with practical skills and knowledge to implement machine learning models using Python, empowering you to tackle real-world data challenges with confidence.

Who this course is for:

Data Scientist looking to further their machine learning knowledge
Beginner ML Engineers aiming to break into machine learning

Introduction to Machine Learning with Python

What you'll learn

Explore related topics

Course content

Introduction to Machine Learning4 lectures • 34min

Regression Models5 lectures • 25min

Classification Models11 lectures • 53min

Clustering4 lectures • 8min

Requirements

Description

Who this course is for: