Machine Learning Python 1 Day Mini Bootcamp
Machine learning is going to disrupt a lot of industries in the next decade. Whether it be driverless cars,cashierless shops, personal assistant or AI physicians, the effect of machine learning will be pervasive.
This class assumes you don’t have any programming background. However, it is recommended to have a basic understanding in Python. Understanding of Pandas Python Library will help.
Although we would use classical datasets like IRIS, Titanic, etc but you will be scale and use your data for the models learned in the session.
You will know whether to run supervised or unsupervised learning for your data, whether to use classification or regression model, how to handle categorical vs continuous data. After the data is ready you will learn how to split the data and analyze the final results. We will use a lot of images to delineate the topics.
Topics covered:
Supervised vs Unsupervised Learning
Regression vs Classification models
Categorical vs Continuous feature spaces
Python Scikit-learn Library
Modeling Fundamentals: Test-train split, Cross validation(CV), Bias–variance tradeoff, Precision and Recall, Ensemble models
Interpreting Results of Regression and Classification Models
Parameters and Hyper Parameters
Dimension Reduction
SVM
K-Nearest Neighbor
Neural Networks
Projects for the session (Python):
Understanding and Interpreting results of Regression and Logistic Regression using Google Spreadsheets and Python
Calculating R-Square, MSE, Logit manually for enhanced understanding
Understanding features of Popular Datasets: Titanic, Iris and Housing Prices
Running Logistic Regression on Titanic Data Set
Running Regression, Logistic Regression, SVM and Random Forest on Iris Dataset
Post Session Assessment:
Top 20 machine learning interview question
Takeaways:
Run machine learning models on your data using the setup learned in class
Make data ready, choose and configure the correct model for your data
Interpret results of your machine learning algorithm