"Turn data into decisions that move the world."
Data Scientists extract insights from data, build predictive models, and communicate findings to drive business decisions. This track takes you from complete beginner to job-ready Data Scientist.
PHASE 1: Foundations (2-3 months)
Python → SQL → Statistics → EDA
↓
PHASE 2: Core ML (3-4 months)
ML Algorithms → Feature Engineering → Model Evaluation → Visualization
↓
PHASE 3: Advanced (4-6 months)
Deep Learning → NLP → Time Series → Experiment Design
↓
PHASE 4: Portfolio & Career
End-to-end Projects → Interview Prep → Job Search
Before starting this track, make sure you have:
- Basic Python programming (loops, functions, data structures)
- Basic SQL (SELECT, JOIN, GROUP BY)
- High school math comfort level (algebra, basic stats)
- A computer with Python 3.10+ installed
Don't have these? Start at Foundations →
| Category | Skills |
|---|---|
| Programming | Python, SQL, R (optional) |
| Data Manipulation | pandas, NumPy, polars |
| Visualization | matplotlib, seaborn, Plotly, Tableau |
| Statistics | Hypothesis testing, regression, Bayesian inference |
| Machine Learning | scikit-learn, XGBoost, LightGBM |
| Deep Learning | PyTorch, TensorFlow, Keras (overview) |
| NLP | NLTK, spaCy, Transformers (basics) |
| Big Data | Spark (basics), SQL on cloud |
| Version Control | Git, DVC |
| Cloud | AWS/GCP/Azure data services (basics) |
- Data storytelling and visualization
- Stakeholder communication (explaining ML to non-technical audiences)
- Problem framing and hypothesis formation
- Experimental design and statistical rigor
- Business acumen and domain knowledge
Goal: Be able to load, clean, analyze, and visualize any dataset confidently.
Duration: 2-3 months (10-15 hrs/week)
- Python for Data Science
- Data manipulation with pandas and NumPy
- Statistics and probability fundamentals
- Exploratory Data Analysis (EDA)
- Data visualization
- SQL for data analysis
| Week | Topic | Resource | Practice |
|---|---|---|---|
| 1-2 | Python for DS refresher | Python for DS | Exercises |
| 3-4 | pandas & NumPy | Data Manipulation Guide | Analyze Titanic dataset |
| 5-6 | Statistics Foundations | Stats Guide | Stats Notebook |
| 7-8 | Data Visualization | Visualization Guide | Create 10 chart types |
| 9-10 | EDA Workflow | EDA Notebook | Full EDA on 2 datasets |
| 11-12 | SQL for Analysis | SQL Guide | SQL Exercises |
Goal: Build, evaluate, and improve ML models for real-world problems.
Duration: 3-4 months (10-15 hrs/week)
- ML fundamentals and theory
- Supervised learning algorithms
- Unsupervised learning
- Feature engineering
- Model evaluation and validation
- Introduction to ensemble methods
| Week | Topic | Resource | Practice |
|---|---|---|---|
| 1-2 | ML Fundamentals | ML Overview | Implement linear regression from scratch |
| 3-4 | Supervised Learning | Classification & Regression | Kaggle House Prices |
| 5-6 | Feature Engineering | Feature Engineering Guide | Engineer 10 new features |
| 7-8 | Model Evaluation | Evaluation Metrics | Cross-validate 5 models |
| 9-10 | Ensemble Methods | Ensembles Guide | XGBoost competition entry |
| 11-12 | Unsupervised Learning | Clustering & Dimensionality | Customer segmentation project |
| 13-14 | ML Pipelines | sklearn Pipelines | Build production-ready pipeline |
Goal: Develop expertise in one or more specialized areas of data science.
Duration: 4-6 months (10-15 hrs/week)
- Deep Learning fundamentals
- Natural Language Processing
- Computer Vision (overview)
- Time Series analysis
- Bayesian Statistics
- Causal Inference and A/B Testing
- Model interpretability (XAI)
| Week | Topic | Resource | Practice |
|---|---|---|---|
| 1-3 | Deep Learning Intro | Deep Learning Guide | Train neural network from scratch |
| 4-6 | NLP Fundamentals | NLP Guide | Sentiment analysis project |
| 7-9 | Time Series | Time Series Guide | Sales forecasting project |
| 10-12 | A/B Testing & Experimentation | Experiment Design Guide | Design and analyze experiment |
| 13-15 | Bayesian Statistics | Bayesian Guide | Bayesian A/B test |
| 16-18 | Model Interpretability | XAI Guide | Explain black-box model |
| 19-21 | Causal Inference | Causality Guide | Propensity score matching |
Portfolio projects are what get you hired. Build at least 3-5 of these:
- Titanic Survival Analysis (EDA + classification)
- Movie Recommendation Analysis (collaborative filtering)
- COVID-19 Dashboard (data viz + storytelling)
- House Price Prediction (full ML pipeline)
- Customer Churn Prediction (classification + business framing)
- Credit Card Fraud Detection (imbalanced classification)
- E-commerce Customer Segmentation (clustering + insights)
- NLP: News Category Classifier with BERT
- Time Series: Stock Price Forecasting with LSTM
- A/B Test: Marketing Campaign Analysis with Causal Inference
- End-to-end: Choose a domain you care about + a real dataset
Use this to track your progress. Fork this repo and check off as you complete each item.
- Can load and clean a messy CSV file with pandas
- Can write SQL queries with JOINs and aggregations
- Understand mean, median, standard deviation, variance
- Can create 5+ different chart types (line, bar, scatter, heatmap, histogram)
- Can perform a basic EDA and write a summary
- Understand normal distribution, central limit theorem
- Can explain p-value to a non-technical person
- Have completed 1 end-to-end EDA project
- Can implement linear and logistic regression
- Understand bias-variance tradeoff
- Can tune hyperparameters with cross-validation
- Know when to use which algorithm (decision tree vs SVM vs XGBoost)
- Can engineer features from raw data
- Understand precision, recall, F1, AUC-ROC
- Have competed in a Kaggle competition
- Can build an sklearn Pipeline
- Have completed 2 ML projects
- Understand backpropagation and gradient descent
- Can fine-tune a pre-trained NLP model
- Can build and evaluate time series models
- Understand SHAP values and model explainability
- Can design and analyze A/B tests statistically
- Understand Bayesian inference
- Have 3+ portfolio projects on GitHub
- Have written a technical blog post about a DS topic
| Book | Author | Link |
|---|---|---|
| Python Data Science Handbook | Jake VanderPlas | Free on GitHub |
| Hands-on ML with Scikit-Learn, Keras, TF | Aurélien Géron | GitHub notebooks |
| Dive into Deep Learning | Zhang et al. | d2l.ai |
| Think Stats | Allen Downey | Green Tea Press (free) |
| The Elements of Statistical Learning | Hastie et al. | Free PDF (official) |
| Course | Platform | Duration |
|---|---|---|
| Machine Learning Specialization | Coursera (Andrew Ng) | 3 months |
| Data Science for Beginners | Microsoft GitHub | 10 weeks |
| Practical Deep Learning for Coders | fast.ai | 14 weeks |
| Kaggle Learn | Kaggle | Self-paced |
| Statistics and Data Science MicroMasters | edX (MIT) | 1 year part-time |
- Kaggle — Competitions and datasets
- LeetCode — SQL and Python practice
- Mode Analytics — SQL for analysis
- StrataScratch — DS interview questions
- Microsoft Certified: Azure Data Scientist Associate
- Google Advanced Data Analytics Professional Certificate
- IBM Data Science Professional Certificate
- AWS Certified Machine Learning — Specialty (if working cloud-side)
- DS Interview Questions →
- Focus areas: Statistics, ML algorithms, SQL, case studies, coding
Back to: Main README | Role Comparison