2-Week Data Science Crash Training¶

A focused, hands-on data science learning vault — structured as a two-week bootcamp — built for students who want to move from Python basics to practical machine learning workflows without getting lost in theory fog.

Trainer: Nikhil Sharma — Sr. Data Science Trainer & Practitioner

Browse the course online at ds.learnwithnickk.com — no setup required.

Week 01 builds the working foundation: Python, NumPy, Pandas, visualization, statistics, SQL, and EDA. Week 02 moves into machine learning, feature engineering, evaluation, deep learning, NLP, and an end-to-end project.

What You Will Learn¶

By the end of this training, you should be able to:

write clean Python for data work
manipulate arrays with NumPy
clean, filter, merge, and analyze datasets with Pandas
create clear charts with Matplotlib and Seaborn
understand core statistics used in data science
write SQL queries for analysis
perform practical exploratory data analysis
train and evaluate beginner ML models
explain your work in interview-friendly language

For Students — Getting Started¶

Two ways to use this course:

Option 1 — Read online (no setup): Open the course website and follow the left-hand navigation. Everything is searchable.

Option 2 — Clone and study locally:

git clone https://github.com/KirkYagami/2-Week-Data-Science-Crash-Training.git
cd 2-Week-Data-Science-Crash-Training

Then open the markdown files in any editor (VS Code recommended).

How to Use This Repo¶

Use the notes in order. The course is intentionally layered, so each section quietly prepares you for the next one.

Recommended rhythm:

Read the note once.
Type the examples yourself.
Do the practice tasks without looking at the solution first.
Keep a small notebook of mistakes and fixes.
At the end of each day, answer the interview questions out loud.

The goal is not to "finish files." The goal is to build reflexes.

Course Roadmap¶

Week 01 — Data Science Foundations¶

Day	Focus	Start Here
Day 01 Part 1	Python Basics	Agenda
Day 01 Part 2	Advanced Python	Agenda
Day 02 Part 1	NumPy Fundamentals	Agenda
Day 02 Part 2	Pandas Basics	Agenda
Day 03 Part 1	Pandas Advanced	GroupBy
Day 03 Part 2	Data Visualization	Matplotlib Basics
Day 04 Part 1	Statistics Basics	Mean, Median, Mode
Day 04 Part 2	Inferential Statistics	Hypothesis Testing
Day 05 Part 1	SQL for Data Science	SELECT and WHERE
Day 05 Part 2	EDA	Data Cleaning

Week 02 — Machine Learning and Projects¶

Day	Focus	Notes
Day 01 Part 1	Machine Learning Basics	Agenda
Day 01 Part 2	Regression Algorithms	Agenda
Day 02 Part 1	Classification Algorithms	Agenda
Day 02 Part 2	Clustering Techniques	Agenda
Day 03 Part 1	Feature Engineering	Agenda
Day 03 Part 2	Model Evaluation	Agenda
Day 04 Part 1	Intro to Deep Learning	Agenda
Day 04 Part 2	NLP Basics	Agenda
Day 05 Part 1	End-to-End Mini Project	Agenda
Day 05 Part 2	Mock Interview and Resume Review	Agenda

Fast-Access Folders¶

External Resources Worth Keeping Open¶

Official Documentation¶

These are the best references when you want the authoritative answer:

Practice Platforms¶

Kaggle Learn — short, practical micro-courses.
SQLBolt — excellent browser-based SQL drills.
Mode SQL Tutorial — SQL for analytics-style questions.
Kaggle Datasets — practice with real-world messy data.

Videos That Actually Help¶

StatQuest Video Index — statistics and ML explained clearly.
3Blue1Brown Neural Networks Playlist — visual intuition for neural networks.
freeCodeCamp Data Analysis with Python — NumPy, Pandas, Matplotlib, and Seaborn in one long practical course.
Stanford CS229 — more theoretical machine learning, useful after the basics.

Papers and Books Worth Knowing¶

You do not need to read all of these during the crash course. Treat them as landmarks: the kind of material you revisit as your understanding grows.

Friendly Books¶

An Introduction to Statistical Learning — one of the best bridges from beginner ML to serious statistical learning.
ISLP Python Companion — Python support for Introduction to Statistical Learning.
Deep Learning Book — a deep reference for neural networks and representation learning.

Classic and Practical Papers¶

Scikit-learn: Machine Learning in Python — the paper behind the library used everywhere in practical ML.
XGBoost: A Scalable Tree Boosting System — important for tabular ML and Kaggle-style modeling.
Adam: A Method for Stochastic Optimization — the optimizer behind many deep learning workflows.
Attention Is All You Need — the Transformer paper behind modern language models.
The Elements of Statistical Learning — more advanced than ISL, but a classic reference.

Suggested Daily Workflow¶

Morning      Read notes + type examples
Midday       Solve practice problems
Afternoon    Build one small analysis from scratch
Evening      Review interview questions + summarize mistakes

For each topic, try to produce one small artifact:

a Python script
a cleaned CSV
a chart
a SQL query
a short EDA report
a model evaluation summary

Small finished artifacts beat large unfinished intentions.

Project Ideas After the Course¶

If you want to keep going, build one of these end to end:

customer churn prediction
house price prediction
sales forecasting
movie recommendation system
sentiment analysis on tweets
Titanic survival prediction

The Projects folder is set up for these.

Quality Checklist¶

When studying or building a project, ask:

Did I inspect the raw data before cleaning?
Did I document missing values and outliers?
Did I avoid data leakage?
Did I choose the right metric?
Did I explain the result in plain English?
Could I defend this in an interview?

Current Status¶

Both weeks are fully written and structured. Week 01 covers the complete data science foundation. Week 02 covers machine learning, deep learning, NLP, and an end-to-end project with mock interview preparation. The cheat sheets, interview prep, and guided projects are available and actively maintained.

Final Note¶

Data science is not one skill. It is a stack of habits: curiosity, cleaning, checking assumptions, writing code that survives tomorrow, and explaining what you found without making the chart do magic tricks.

Use this repo as a training ground. Break things, fix them, write the insight down, and keep moving.