Algorithmic Introduction to Machine Learning

Name: Algorithmic Introduction to Machine Learning
Rating: 4.4 (5 reviews)

A Conceptual Framework

Created byNisheeth Joshi

Last updated 2/2023

English

English [Auto],

What you'll learn

Understand the process of preprocessing data
Understand the working behind various machine learning algorithms
Use different evaluation measures and decode confusion matrix
Use different machine learning techniques to design AI machine and enveloping applications for real world problems.

Course content

5 sections • 16 lectures • 7h 36m total length

Course Introduction2:24
Gain a practical overview of machine learning, covering data pre-processing, missing values, encoding categorical data, and normalization, and explore how supervised and unsupervised algorithms work with cross-validation.
Intro to Machine Learning47:46
Understand what machine learning is and how data-driven learning produces predictions through supervised, unsupervised, and reinforcement learning, with a five-step workflow from data gathering to testing.

Handling Missing Values43:53
Explore data pre-processing by addressing missing values to build robust machine learning models, covering deletion, mean/median imputation, unique category encoding for categorical data, predictive imputation, and methods like k-nearest neighbors.
Data Encoding33:10
Learn to convert categorical data into numeric formats through encoding, including binary encoding and probability-based and average-based target encoding, enabling regression, SVMs, and neural networks.
Data Normalization38:27
Learn range normalization, or scaling, to map features like age and salary into a common range such as -1 to 1 or 0 to 1, and apply denormalization to revert.

Linear Regression - I43:24
Learn linear regression as predictive modeling, using simple and multiple regression to relate independent and dependent variables, and apply regression lines for forecasting and risk assessment with examples.
Linear Regression - II26:44
Explore linear regression concepts from y equals mx plus c to multiple regression, minimize error with gradient descent, assess fit with r-squared, and preview manual and scikit-learn implementations.
Decision Tree Regression53:02
Construct a regression decision tree by splitting on attributes with the highest standard deviation reduction to predict a continuous target, such as hours played.
Decision Tree Classification34:54
Learn how decision trees classify data using a top-down greedy ID3 approach, building decision and leaf nodes, calculating entropy and information gain to split datasets.
Naive Bayes Classification12:54
Explore the naive Bayes classifier, based on Bayes theorem, using prior, likelihood and Laplace smoothing to compute posteriors under conditional independence of features.
KNN Classification14:25
K-nearest neighbor classification assigns a point to the class of its nearest neighbors using Euclidean, Manhattan, or Minkowski distances, with odd k (3–9) and data standardization to prevent scale bias.
Model Evaluation15:25
Learn how to evaluate classification models using confusion matrices, calculate true/false positives and negatives, and derive accuracy, precision, recall, and f-measure for binary and multi-class tasks.

Clustering Introduction7:36
Explore clustering as an unsupervised learning method that forms clusters of similar data using distance measures like euclidean, manhattan, and minkowski, with partitive and hierarchical algorithms.
KMeans Clustering12:32
K-means clustering partitions data into k clusters by assigning points to the nearest centroid and updating centroids as means until convergence; the optimal k is data-driven.
Hierarchical Clustering33:08
This lecture explains hierarchical clustering, showing divisive and agglomerative approaches that build a cluster hierarchy by top-down division or bottom-up merging using distance measures.

Requirements

Basic knowledge of data structures is required

Description

Machine learning (ML) is a branch of artificial intelligence (AI) that enables computers to “self-learn” from training data and improve over time, without being explicitly programmed. Machine learning algorithms are able to detect patterns in data and learn from them, in order to make their own predictions. This course is intended for people who wish to understand the functioning of popular machine learning algorithms. This gives a behind the scene look of who things are working. We will start by looking at some data pre-processing techniques, then we will move on to look at supervised and unsupervised learning algorithms. Finally, we will look at what cross valuation is and how it is done.

In this course we will look at: Data Preprocessing [Handling Missing Values, Data Encoding (Conversion of Categorical Data into Nominal Data), Data Normalization] Supervised Learning[Linear Regression, Decision Tree Regression, Decision Tree Classification, Naive Bayes Classification, K Nearest Neignbour Classification] Model Evaluation [Evaluation of Classifiers, Deciding Confusion Matrix] Unsupervised Learning [K Means Clustering, Hierarchical Clustering] Model Improvement [Cross Validation]

By the end of this course, you will have a thorough understanding of how these machine learning algorithms function which will in turn enable you to develop better ML models.

Who this course is for:

Beginners desirous of understanding how machine learning algorithms work

Algorithmic Introduction to Machine Learning

What you'll learn

Explore related topics

Course content

Introduction2 lectures • 50min

Data Preprocessing3 lectures • 1hr 56min

Supervised Learning7 lectures • 3hr 21min

Unsupervised Learning3 lectures • 53min

Cross Validation1 lecture • 37min

Requirements

Description

Who this course is for: