Python for Mastering Machine Learning and Data Science

Master Python basics from the starter kit: variables, data structures (lists, tuples, sets, dictionaries), loops, conditionals, functions, comprehensions, lambdas, built-ins, and external libraries, with hands-on exercises and solutions.

Learn how to define variables in Python, assign numbers and text, and print them. Explore types like string, integer, float, and boolean, and use type to check variable types.

Explore Python string basics with zero-based and negative indexing, and slicing using examples like a name and color. Apply string methods such as lower, upper, and title to transform text.

Learn how to replace hardcoded names with variables using formatted strings (f-strings) and curly braces, so changing one variable updates all appearances; variables store strings and enable reuse across code.

Explore Python lists by creating, appending, and removing elements using methods like append, pop, and insert. Learn indexing, slicing, and mutability to manage mixed-type data.

Explore tuples in Python, compare them with lists, and learn to create, access, slice, and convert between tuples and lists while understanding their immutability.

Explore how the while loop repeats actions based on a true condition, contrasts with the for loop's iterations, and demonstrates an ammo count example that ends when ammo reaches zero.

Learn how to define and use custom Python functions to perform tasks, pass arguments, handle input, and return results, with examples like len and summing lists.

Explore Python exercises spanning two parts, from user input and string manipulation to list, dictionary, and function tasks, emphasizing basic operations and introductory problem solving.

Practice Python exercises on user input, string to list conversion, rounding, indexing, and absolute values. Also cover odd/even checks, vowel masking, string splitting, deduplication with sets, and dictionary updates.

Explore linear regression theory, from simple to multiple regression, and learn to fit a regression line with the least squares method to predict the dependent variable from the independent variable.

Import and rename pandas, numpy, matplotlib, and seaborn to pd, np, plt, and sns. Load accountant salary data with pd.read_csv, and explore with head, info, describe, corr, and a heatmap.

Define X as all features except property value and y as value, then perform a train-test split. Scale after splitting, train linear regression model, and evaluate with MAE, MSE, RMSE.

Logistic regression uses a sigmoid function for binary classification and relies on maximum likelihood, predicting probabilities and class labels such as obese versus not obese.

Study logistic regression for gender classification, covering data import, EDA with heat map, missing values, and feature engineering, then train the model and evaluate with confusion matrix and classification report.

Explore logistic regression for gender classification by loading a made-up gender classification dataset, inspecting data quality, handling missing values, and preparing features like hair length for model readiness.

Learn how linear regression is enhanced by regularization through ridge and lasso, using lambda penalties to shrink coefficients, reduce variance, and for lasso, exclude irrelevant features.

Master L1 and L2 regularization with lasso and ridge regression on the Melbourne housing dataset, focusing on data cleaning, handling missing values, and creating dummy variables.

Clean and engineer Melbourne housing data by dropping unused features and one-hot encoding categoricals. Convert date to datetime, extract year and month, and prep features for lasso and ridge regression.

Apply ridge and lasso regression to predict insurance charges from a small dataset. Convert sex and smoker features to numerical values and create region dummy variables, then compare models.

Address imbalanced data in classification by balancing classes with under-sampling, over-sampling, and synthetic minority sampling techniques, and evaluate with multiple metrics to avoid the accuracy paradox.

Detect and remove outliers using box plots, z-scores, and the IQR in a pandas dataframe, with practical steps and visualization.

Explore feature scaling with min‑max, standard, and robust scalers, compare their effects on linear regression, and evaluate using mean absolute error and root mean squared error after a train‑test split.

Apply Gaussian naive bayes to Titanic data, perform exploratory data analysis, drop cabins with many missing values, and impute age by passenger class means before building a survivor predictor.

Prepare the Titanic dataset by cleaning, encoding categorical features with dummies, and selecting features for a Gaussian Naive Bayes classifier. Train and evaluate with confusion matrix and classification report.

Explore how a decision tree uses root and leaf nodes to make predictions, handles numerical and categorical features, and selects root feature by impurity measures like Gini impurity or entropy.

Master a decision tree regression on a wine quality dataset to predict quality. Load data, split into train and test, train the model, and evaluate with MAE, MSE, and RMSE.

Clean and prepare bike sharing data with dummy variables and log normalization of the count. Train a random forest regressor and apply grid search tuning to improve performance.

Build a random forest to predict rain tomorrow in Australia, perform preprocessing, handling missing values, and create dummy variables. Evaluate and address class imbalance with SMOTE and grid search CV.

Clean weatherAUS data, visualize correlations with a heat map, create dummy variables for categorical features, and train a random forest classifier while addressing imbalance and tuning with grid search.