Comprehensive Study Guide: Python, Data Analytics & Machine Learning

These flashcards cover key concepts, terms, and definitions related to Python, Data Analytics, and Machine Learning as outlined in the comprehensive study guide.

Data Types

int, float, string, boolean.

List Characteristics

Ordered, mutable, allow duplicates, index starts at 0.

Dictionaries

Key–value pairs used in Python.

Function Parameters vs Arguments

Parameters are in function definition; arguments are values passed.

Python Case Sensitivity

Python is case sensitive.

Negative Indexing

Counts from the end of a list.

df.head(n) function

Returns first n rows of a DataFrame.

df.iloc[:n] function

Returns rows from index 0 to n-1.

DataFrame Mean Calculation

df['column'].mean() calculates the average.

Pandas Import

Pandas is NOT included in base Python and must be imported.

Supervised Learning

Involves labeled data for classification and regression.

Unsupervised Learning

Involves unlabeled data for clustering and dimensionality reduction.

Binary Classification

Predicts between two classes.

Multi-class Classification

Involves more than two mutually exclusive classes.

Multi-label Classification

Allows multiple labels per observation.

Classification Predictions

Predicts categories, not continuous values.

Accuracy in Model Evaluation

Measures overall correctness; weak for imbalanced data.

Precision

Correctness of positive predictions.

Recall

Ability to identify actual positives.

F1-Score

Balance of precision and recall.

R² in Regression

Proportion of variance explained.

RMSE

Penalizes large errors; same units as the target.

MAE

Average absolute error; less sensitive to outliers.

KNN

A supervised learning method relying on Euclidean distance.

K-Means Clustering

An unsupervised clustering method reliant on Euclidean distance.

Elbow Method & Silhouette Score

Help determine the optimal number of clusters (K).

Time Series Analysis Purpose

Used to forecast values over time.

Time Series Components

Trend, Seasonality, Residual (Irregular).

Additive Model in Time Series

Seasonal magnitude is constant.

Multiplicative Model in Time Series

Seasonal magnitude is proportional to trend.

Forecasting Requirements

Requires lagged values and time-based predictors.

R² and Forecasting

High R² does NOT guarantee good forecasts.

CRISP-DM Framework

Steps include Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation, Deployment.

Four V's of Big Data

Volume, Velocity, Variety, Veracity.

Visualization Purposes

Used for exploratory analysis and storytelling.

Exam Strategy Tips

Match business questions to modeling approaches; precise terminology; watch for default behaviors in Python and Pandas.