Complete Machine Learning Course in 60 Hours - Part 3 | Full Machine Learning Course for Beginners | Siddhardhan YouTube Video Summary
Siddhardhan introduces part three of his complete 60-hour machine-learning course. He recaps Part 1, which covered machine learning basics, Python libraries, and data pre-processing. Part 2 explored mathematics for machine learning and machine learning models, including linear algebra, statistics, probability, and gradient descent. Part 3 focuses on building machine learning models from scratch, including linear regression, logistic regression, support vector machines, and lasso regression.
Detailed Summary:
Complete Machine Learning Course in 60 Hours - Part 3
Siddhardhan introduces part 3 of his complete machine-learning course in 60 hours.
Recap of Part 1
Part 1 covered Machine Learning – Basics, Python basics for Machine Learning, Python Libraries, Data Collection & Pre-Processing, and a number of use cases:
- Rock vs Mine Prediction
- Diabetes Prediction
- Spam Mail Prediction
Recap of Part 2
Mathematics for Machine Learning (Linear Algebra, Statistics, and Probability), Machine Learning Models, Working of ML models, Model Selection, Overfitting and Underfitting, Loss Function & Model Evaluation, Model Parameters & Hyperparameters, Gradient Descent and the following use cases:
- Heart Disease Prediction
- Boston House Price Prediction
- Loan Approval Prediction
Part 3: Building Machine Learning Models From Scratch
Part 3 has one module that covers the following:
- Linear Regression
- Logistic Regression
- Support Vector Machine
- Lasso Regression
Each model will have videos on: Intuition of Model, Math Behind Model, Build Model from Scratch and How to Implement in Python.
Understanding How a Model Works
A machine learning model consists of data that you feed into the machine for the model to learn. Once it learns from the data it tries to understand the patterns that are present, then if you give a new set of data it is able to predict the new data based on the patterns that have been understood. Loss function measures how far an estimated value is from its true value. It is helpful to determine which model performs better & which parameters are better.
Linear Regression Explained
A linear regression is a model you use with 2 sets of data: x axis and y axis, The linear regression model tries to fit the data set to a straight line, that may or may not be in the middle of all data points. It uses 2 values - an intercept and a slope. You find these with an equation y=mx+c. The goal is to get the model the point of understanding that there is a linear relationship and with the right slope, it knows if x values increase, the y values are probably going to increase as well. However linear regression is not suitable for non-linear datatypes like a sine wave, it also has a number of issues if there are too many outliers in your data.