Data Science and Machine Learning in Python

Learn Python basics fast by using Python as a calculator, mastering variables, string operations, and zero-based indexing with slicing and common string methods.

Explore Python lists, tuples, and dictionaries: create and index lists, mutate elements, understand nested structures, grasp tuple immutability, and map keys to values for JSON-like data.

Practice practical Python exercises and solutions, covering arithmetic, string operations, and formatting. Learn indexing in lists and dictionaries, domain extraction from emails, and a conditional speed ticket function.

Explore NumPy operations part 3, applying basic inbuilt functions like min, max, argmax, and algebraic operations such as multiply, add, and divide on arrays, with notes on transpose and sorting.

Learn how Pandas reads data from various formats, selects and filters data, sorts and groups, renders plots, and computes descriptive statistics like mean and variance.

Explore pandas basics: import libraries, load csv data from a bank marketing data set, and inspect head, tail, shape, columns, and dtypes to prepare data for machine learning.

Master Pandas row selection and slicing with iloc, accessing the fifth or last row, and combining row and column lists for targeted data extraction from a bank loan data set.

Learn to compute descriptive statistics in pandas: select columns, perform numeric operations, use describe to summarize central tendency and dispersion, and inspect data types and missing values.

Learn how to handle missing data in pandas: drop or fill missing values, impute with column means or carrier-based group means, and apply cleaning to prepare data for machine learning.

Learn to use pandas groupby and sorting to compute mean, sum, max, min, and median by job or education, then reset indices for clear, organized data insights.

Learn how to use pandas pivot tables and cross tabulations to generate frequency and summary statistics, including mean, sum, and median, from data, exploring education, marital status, and job categories.

Master time series data in pandas by extracting date time attributes, creating date ranges with time zones, and filling missing values using interpolate, forward fill, or backfill.

Explore pandas techniques for merging, joining, and concatenating dataframes, using axis options and on keys to perform sql-style joins.

Import and export data in Python with pandas, reading csv, excel, json, html, clipboard, AWS S3, and SQL databases; use read_* methods and SQLAlchemy for end-to-end data pipelines.

Visualize data distributions in Python using seaborn, distplot with kde and rug options, and customize bins to compare normal and gamma distributions; explore Iris data with grids for multivariate patterns.

Explore plotting categorical variables with Seaborn in Python, using swarm, box, violin, bar, count, and point plots across tips, diamonds, and Titanic datasets, with hue and other aesthetics.

Explore visualizing statistical relationships using scatter and line plots with Seaborn, including hue, size, and style, on tips and time-series data.

Explore scikit-learn's algorithms for machine learning, covering supervised learning (classification and regression), unsupervised learning (clustering), start with linear regression, and address overfitting and underfitting.

Master data preprocessing in python, covering standardization and normalization, scaling methods, and encoding categorical features with one-hot encoding. Explore discrete bucketing and polynomial features for feature engineering.

Split data into training and testing sets using train_test_split, choose a test_size (for example 0.33) and set random_state for reproducibility, and consider the 80 20 rule depending on data size.

This lecture introduces a practical supervised learning template in python, covering data import, exploratory analysis, null handling, get dummies for categoricals, train-test split, scaling, model training, and evaluation.

Explore practical linear regression in Python: load data, preprocess, split training and test sets, fit the model, predict, and interpret coefficients with evaluation metrics on the Boston housing dataset.

Master linear regression through a guided exercise that walks you through steps to build a regression model, learn procedures, and apply the method in practice.

Explore comparing multiple regressors—linear, ridge, lasso, decision tree, k-nearest neighbors, gaussian process, and random forest—on air quality data, evaluated with R2 score, explained variance, and MAE.

Explore decision tree and random forest classifiers on wine quality data, and master model selection with grid search cross-validation to optimize parameters.

Apply support vector machines to classify breast cancer data, use grid search cross-validation to optimize C and gamma, and evaluate improvements with confusion matrix and classification report.

Apply multiple classifiers to the car dataset by reading the csv file from the folder. Reference your earlier notebooks as needed and prepare for the solution lecture.

Learn a practical python workflow for classification: import, inspect, encode categoricals with get_dummies, and train-test split. Compare multiple classifiers and select the top model for a car dataset.

Explore k-means clustering in python, using make blobs to generate toy data, identify cluster centers and labels, and visualize unsupervised grouping in practice.

Explore model selection and hyperparameter tuning with grid search cross-validation on the Wisconsin breast cancer dataset with 30 features, comparing k-nearest neighbors using distance metrics to improve accuracy.

Explore feature engineering and dimensionality reduction techniques, including creating new features, PCA, ICA, NMF, LDA, and feature selection, to improve model performance on datasets like breast cancer and digits.