Machine Learning for Data Analysis: Regression & Forecasting

HEADS UP. This course is now part of The Complete Visual Guide to Machine Learning & Data Science, which combines all 4 Machine Learning courses from Maven Analytics. This course, along with the other individual courses in the series, will be retired soon.

This course is PART 3 of a 4-PART SERIES designed to help you build a strong, foundational understanding of Machine Learning:PART 1: QA & Data ProfilingPART 2: Classification ModelingPART 3: Regression & ForecastingPART 4: Unsupervised LearningThis course makes data science approachable to everyday people, and is designed to demystify powerful Machine Learning tools & techniques without trying to teach you a coding language at the same time. Instead, we'll use familiar, user-friendly tools like Microsoft Excel to break down complex topics and help you understand exactly HOW and WHY machine learning works before you dive into programming languages like Python or R. Unlike most Data Science and Machine Learning courses, you won't write a SINGLE LINE of code.

COURSE OUTLINE:In this Part 3 course, we'll start by introducing core building blocks like linear relationships and least squared error, then show you how these concepts can be applied to univariate, multivariate, and non-linear regression models. From there we'll review common diagnostic metrics like R-squared, mean error, F-significance, and P-Values, along with important concepts like homoscedasticity and multicollinearity. Last but not least we'll dive into time-series forecasting, and explore powerful techniques for identifying seasonality, predicting nonlinear trends, and measuring the impact of key business decisions using intervention analysis:Section 1: Intro to RegressionSupervised Learning landscapeRegression vs.

ClassificationFeature engineeringOverfitting & UnderfittingPrediction vs. Root-Cause AnalysisSection 2: Regression Modeling 101Linear RelationshipsLeast Squared Error (SSE)Univariate RegressionMultivariate RegressionNonlinear TransformationSection 3: Model DiagnosticsR-SquaredMean Error Metrics (MSE, MAE, MAPE)Null HypothesisF-SignificanceT-Values & P-ValuesHomoskedasticityMulticollinearitySection 4: Time-Series ForecastingSeasonalityAuto Correlation Function (ACF)Linear TrendingNon-Linear Models (Gompertz)Intervention AnalysisThroughout the course we'll introduce hands-on case studies to solidify key concepts and tie them back to real world scenarios. You'll see how regression analysis can be used to estimate property prices, forecast seasonal trends, predict sales for a new product launch, and even measure the business impact of a new website design.