Machine Learning 003. Data preprocessing part 2: Handling missing values Data cleaning

Описание: 🎓 Video Description: Lecture 3 – Data Cleaning & Handling Missing Values

Welcome back to our Machine Learning series!
In this video, we dive deeper into Data Preprocessing, focusing on one of the most important steps in building reliable models — Data Cleaning.

We’ll explore how to identify and fix missing, invalid, and inconsistent data that can reduce your model’s accuracy.
This lecture continues from Lecture 2 (Data Preparation & Model Evaluation) — and here, we focus specifically on handling missing values using multiple imputation methods.

🔍 What You’ll Learn:

✅ What Data Cleaning is and why it matters
✅ Common data quality issues (missing, invalid, misfielded, duplicates, format errors)
✅ Why missing data is a major problem for machine learning models
✅ How to handle missing values with:

Mean / Median imputation

Mode (Most Frequent) imputation

Constant / Zero imputation
✅ Pros and cons of each imputation strategy
✅ When to avoid dropping rows and columns
✅ Practical insights for real-world datasets

🧩 Chapters:

00:00 – Intro: Where Data Preprocessing Fits in ML
00:40 – What Is Data Cleaning?
02:00 – Common Data Quality Issues
03:00 – Handling Missing Values
04:15 – Mean & Median Imputation
05:15 – Most Frequent and Constant Value Imputation
07:00 – Summary & What’s Next

💻 Next Lecture (Lecture 4):

👉 Advanced Data Preprocessing – Encoding Categorical Variables & Scaling Features
We’ll explore label encoding, one-hot encoding, feature scaling, and normalization techniques used to prepare data for model training.


🧠 Tags / Keywords:

#MachineLearning #DataScience #DataPreprocessing #DataCleaning #MissingValues #DataImputation #MeanImputation #ModeImputation #FeatureEngineering #MSDA3050 #PythonDataScience