Machine Learning Essentials with Python-From Zero to Mastery

Name: Machine Learning Essentials with Python-From Zero to Mastery
Rating: 4.6 (12 reviews)

K-NN, Linear Regression, SVM, K Means Clustering, Decision Tree, Neural Networks, Deep Learning and Convolutional NNs

Created byDr. Zahra Golrizkhatami

Last updated 12/2025

English

English [Auto],

What you'll learn

You will learn data science, pattern recognition and machine learning all using Python.
Have a great intuition of many Machine Learning models
Implement popular Machine Learning Algorithms such as KNN, SVM, Linear Regression, K Means Clustering and Decision Tree
Know which Machine Learning model to choose for each type of problem

Course content

16 sections • 90 lectures • 11h 8m total length

Course Intro2:19
Demystifying Machine Learning8:17
Machine Learning as Finding a Hidden Rule6:44
This lecture introduces the core concepts of supervised machine learning by framing the field as a form of function approximation. The sources define the mathematical relationship between input features and target labels, illustrating these ideas through medical diagnosis scenarios and a whimsical interview with a preschooler. Students are introduced to fundamental tasks such as classification, where outputs are discrete categories, and regression, which involves real-valued predictions. Key algorithms like majority vote, memorizers, and decision stumps are presented alongside methods for measuring performance using loss functions and error rates. Finally, the text highlights the transition to more complex decision trees, emphasizing that the ultimate goal of a learner is to develop a hypothesis that generalizes well to unseen data.
Machine Learning: Concepts to Classification6:28
Explore how supervised learning uses labeled data to train models and classify new data with the iris data set and k-nearest neighbors.
Python Numpy Tutorial (with Jupyter and Colab)0:55
Machine Learning: The Science of Function Approximation7:24
This presentation defines machine learning as the process of function approximation, where algorithms attempt to replicate unknown real-world patterns using available data. By converting observations into structured features and labels, researchers can apply a universal framework to solve diverse problems ranging from mathematical puzzles to medical diagnoses. The core of this system involves creating a hypothesis that can generalize its predictions to handle variety in classification or regression tasks. To ensure these models are effective, they must be rigorously validated against unseen test data to measure their accuracy. Ultimately, the supervised learning lifecycle relies on utilizing loss functions to quantify errors and refine the predictive performance of the artificial model.
Python and NumPy for Machine Learning6:36
Ensembles, Bagging, and Linear Regression Fundamentals8:44
Explore ensembles and bagging to reduce variance with a random forest example. Build intuition for linear regression with gradient descent, loss functions, and L2 regularization to prevent overfitting.

Regularization, Cross-Validation, and the Bias-Variance Tradeoff8:00
This lecture explores key concepts in machine learning, focusing on model complexity and methods to prevent overfitting. It introduces the Bias-Variance Tradeoff, explaining how complex models typically have low bias but high variance, leading to overfitting. The text then details two primary strategies for choosing optimal model complexity without compromising the test set: using a validation set and employing k-fold cross-validation, outlining the pros and cons of each. Finally, the material introduces regularisation as a technique to mitigate overfitting by penalising large coefficients, specifically discussing Ridge Regression (L2 penalty) and LASSO Regression (L1 penalty), and comparing their characteristics, such as LASSO's tendency to promote sparsity.
Bias-Variance Tradeoff and Cross-Validation6:35
Explore the bias-variance tradeoff in machine learning, distinguishing overfitting and underfitting, and learn validation strategies like holdout and k-fold cross-validation to estimate model performance.
Getting Started1:01
Mean, Median, and Mode1:35
What is Standard Deviation?2:11
What are Percentiles?0:44
Scatter plot0:53

How Machine Learn to Predict?8:30
Explore how supervised learning turns data into predictions by fitting a line for house prices, then extend to classification with logistic regression via the GLM framework.
A Guide to Linear Regression(2025)8:02
Explore linear regression and ordinary least squares to predict outcomes from features, extend to multiple and polynomial regression, and implement a simple five-step workflow with NumPy, Scikit learn, and Statsmodels.
Data Preparation15:58
Master data preparation by reading and trimming a multivariate student performance dataset, then apply a linear regression model to predict final grades from selected attributes.
How Linear Regression Works?7:29
Linear Regression Example: Student Performance Estimation8:12
Saving Model & Plotting Data14:59
Save trained models with pickle and preserve the best by accuracy, then plot final grade against study time, failures, and absences with a scatter plot on a grid.
Assignment 1 with Solution
Resources Related to the Assignment 10:02
Logistic Regression2:28

Foundations of K-Nearest Neighbors Algorithm(2025)6:59
Explore k nearest neighbors, a lazy, non-parametric classifier that votes among the closest neighbors to label a new point. Learn to pick k with cross-validation.
Irregular Data13:59
How K Nearest Neighbors works?9:31
Explore how the k-nearest neighbors algorithm classifies data by voting among the closest points using distance measures such as Euclidean, Manhattan, and Hamming, and trade-offs in training time and memory.
K-NN Implementation10:35
Implement k-nearest neighbors classification by training a knn model, evaluating accuracy on test data, and tuning the number of neighbors.
K-NN Regression Example
K-NN Classification Example
Mastering the Bias-Variance Tradeoff in Prediction7:45
This lecture focuses on the core machine learning concept of the bias-variance tradeoff for model generalization. It explains that a model's total error is a combination of systematic error (Bias), which measures accuracy, and model instability (Variance), which measures consistency, along with irreducible noise. The source illustrates that models can fail by being too simple (underfitting), resulting in high bias, or too complex (overfitting), resulting in high variance, as shown through polynomial regression examples. The ultimate goal is to find the optimal model complexity that minimizes total error by balancing these two components. Finally, the text introduces Regularization and Cross-Validation as practical techniques used to manage model complexity and achieve the necessary balance for accurate prediction on new data.

SVM Foundations(2025)7:48
Support Vector Machines: Lines, Margins, and Kernels8:11
Discover how support vector machines classify data by maximizing the margin with a line, guided by support vectors, and extend to non-linear separation via the kernel trick and soft margins.
Using Sklearn Datasets5:32
How SVM Works?8:05
SVM Implementation8:29
Implement a support vector classification model, explore linear and other kernels, and tune the C parameter to improve accuracy, while highlighting supervised learning and upcoming K means clustering.

K-Means Clustering: Definition, Mechanics, and Optimization7:07
Explore how k-means clustering, an unsupervised centroid-based algorithm, partitions data into k hard clusters, iteratively assigns points and updates centers until convergence, with elbow method and k-means++ for better results.
How K Means Clustering Works?13:56
Learn how k-means clustering uses unlabeled features to form clusters by assigning points to the nearest centroid and updating centroids to the mean, iterating until convergence, illustrated with digit classification.
K Means Clustering Implementation10:08
Learn how to implement k-means on the digits dataset, scale features for fair euclidean distance calculations, and evaluate with homogeneity, completeness, adjusted Rand, mutual information, and silhouette scores.
Assignment with Solution
Resources related to the K-Means Assignment0:02
Hierarchical and Spectral Clustering Techniques7:11
Explore how agglomerative clustering builds a dendrogram from single points to many groups, and how spectral clustering uses eigenvectors and normalized cuts to reveal nonconvex structures.

Decision Trees: The AI You Use Every Day (2025)8:30
Explore decision trees as simple yet powerful models, using information gain to split data, building transparent trees with nodes and leaves, and preventing overfitting through pruning.
How Decision Tree Works?5:34
Discover how a decision tree, a supervised learning method, builds splits using information gain, reduces overfitting with pruning and max depth, and contrasts with random forests.
Decision Tree Implementation13:21
Learn how to implement a decision tree classifier on the breast cancer dataset, including preparing features, training the model, making predictions, and evaluating accuracy, recall, and a confusion matrix.
Decision Tree Example0:03
DT Assignment with Solution
Deconstructing Decision Trees-20257:24
Explore how a decision tree splits data with internal questions, uses entropy and information gain to select the best splits, and employs pruning to prevent overfitting.
Downloadable resource for DT Assignment + Solution0:02

Ensemble Learning6:59
Explore how ensemble learning turns simple models into a powerful predictor by overcoming bias-variance trade-off through bagging and boosting, as seen in the Netflix prize.
Lone Genius vs. The Crowd7:12
Ensemble Learning: Collective Intelligence, Bagging, and Boosting6:27
This lecture offers an introduction to ensemble learning, which involves combining multiple predictive models to achieve a stronger final result, referencing the Netflix recommendation challenge as a prime example of this strategy. A significant portion of the slides explains the Bias-Variance Tradeoff in predictive error and explores the first ensemble method, Bagging (Bootstrap Aggregation), as a technique to reduce variance and combat overfitting by averaging independent estimates from resampled datasets. Subsequently, the source examines Boosting as an alternative sequential ensemble approach designed to reduce bias by training models to correct the mistakes of previous weak learners, detailing the AdaBoost algorithm and its theoretical foundation in minimizing exponential loss. Ultimately, the material contrasts Bagging and Boosting, summarizing the former as a parallel method for variance reduction and the latter as a sequential method for bias reduction.

Text Classification: Naive Bayes, Evaluation, and Harms8:54
Explore how text classification, via Naive Bayes and bag-of-words, learns to distinguish positive and negative text, evaluates with precision and recall, and addresses harms and bias.
Naive Bayes for Document Classification and Smoothing7:35
Learn how Naive Bayes classifies documents with a bag-of-words model, using prior probabilities and Laplace smoothing to avoid zero probabilities, while handling underflow with logarithms.

Requirements

Basic knowledge of Python programming
You'll need a desktop computer (Windows, Mac, or Linux) capable of running Anaconda 3 or newer. The course will walk you through installing the necessary free software.

Description

This course contains the use of artificial intelligence. Some of the videos in this course were created using AI-assisted tools. These tools were used to professionally produce high-quality visuals and narration in order to make the learning process clearer, more engaging, and more efficient. All learning materials were carefully selected, organized, and updated by the instructor to reflect current knowledge and best practices. AI was used as a supportive technology, not as a substitute for subject-matter expertise, instructional design, or academic responsibility.

Update(02/12/2025): Tens of NEW Lecture Videos and Jupiter Notebooks have been added.

Are you interested in the field of machine learning? Then you have come to the right place, and this course is exactly what you need!

In this course, you will learn the basics of various popular machine learning approaches through several practical examples. Various machine learning algorithms, such as K-NN, Linear Regression, SVM, K-Means Clustering, Decision Trees, Hidden Markov Models and Reinforcement Learning, Bayesian Networks, Neural Networks, Deep Learning and Convolutional Neural Networks, will be explained and implemented in Python. In this course, I aim to share my knowledge and teach you the basics of the theories, algorithms, and programming libraries in a straightforward manner. I will guide you step by step on your journey into the world of machine learning.

Each concept is introduced in plain English, avoiding confusing mathematical notation and jargon. It’s then demonstrated using Python code you can experiment with and build upon, along with notes you can keep for future reference. You won't find academic, deeply mathematical coverage of these algorithms in this course - the focus is on practical understanding and application of them. This course will teach you the basic techniques used by real-world industry data scientists. I'll cover the fundamentals of machine learning techniques that are essential for real-world problems, including:

Linear Regression
K-Nearest Neighbor
Support Vector Machines
K-Means Clustering
Decision Tree
Markov Models and Reinforcement Learning
Bayesian Networks,
Neural Networks
Deep Learning
Convolutional Neural Networks

These are the basic topics any successful technologist absolutely needs to know about, so what are you waiting for? Enrol now!

Who this course is for:

Beginner Python developers curious about data science

Machine Learning Essentials with Python-From Zero to Mastery

What you'll learn

Explore related topics

Course content

Introduction8 lectures • 47min

Introduction and Setup1 lecture • 12min

Introduction to Statistics for Machine Learning7 lectures • 21min

Linear Regression8 lectures • 1hr 6min

K Nearest Neighbors Classification7 lectures • 49min

Support Vector Machine5 lectures • 38min

K Means Clustering5 lectures • 38min

Decision Tree6 lectures • 35min

Ensembling, Linear Models, and Optimization Principles3 lectures • 21min

Naive Bayes2 lectures • 16min

Requirements

Description

Who this course is for: