Data Science & Machine Learning
An advanced, job-oriented course in Data Science and Machine Learning with Python. You wrangle and explore data with Pandas, NumPy and Matplotlib, learn the statistics behind decisions, and build and tune Scikit-learn models without data leakage. You then go further into unsupervised learning, neural networks with PyTorch, NLP and model deployment basics. You use GitHub Copilot throughout and learn to review what it writes. You finish with four portfolio-style projects, an integrated capstone and interview preparation. All business data is simulated. You need Python basics and basic statistics before you start.
- Duration: 32 weeks
- Modules: 12
- Fee: Rs 24,999 (about Rs 4,166 a month on EMI)
- Field: Data & Analytics
Roles this track prepares you for
- Junior Data Scientist
- Associate Data Scientist
- Data Scientist
- Machine Learning Engineer
- Junior Machine Learning Engineer
- AI Engineer
- Data Analyst
- NLP Engineer
- Applied Machine Learning Engineer
Curriculum
- Data Science Workflow, Python Readiness and Copilot Lab Setup - Learn how a data science project runs from question to result, check your Python readiness, set up a reproducible Jupyter environment and learn to use and review GitHub Copilot.
- The data science workflow from question to result
- Python readiness check and gap fixing
- Jupyter and a reproducible Python environment
- Git and GitHub for notebooks and projects
- Set up GitHub Copilot and write prompts you can check
- Review, test and reject AI-written code
- NumPy, Pandas and Data Wrangling - Work with arrays and DataFrames, select and group data, merge tables, reshape data and handle missing values, duplicates and wrong types, and use Copilot carefully while you do it.
- NumPy arrays and vectorized thinking
- Pandas Series and DataFrames
- Select, Filter and Groupby
- Merge, Join and Reshape
- Missing Data, Duplicates and Types
- Copilot-Assisted Wrangling
- Exploratory Data Analysis and Visualization with Matplotlib - Ask questions of the data before modeling, draw clear and honest charts with Matplotlib, find outliers and relationships, and write an exploratory report a manager can read.
- Start EDA with business questions
- Matplotlib figures, axes and clear labelling
- Distributions, outliers and summary statistics
- Relationships, correlation and what it does not prove
- Time and category patterns
- Write an EDA report notebook
- Statistics, Probability and Experiment Thinking - Learn the statistics behind data science decisions: distributions, sampling, confidence intervals, hypothesis tests and A/B tests, and learn to judge whether a result can be trusted.
- Descriptive statistics and common distributions
- Probability, sampling and the central limit idea
- Confidence intervals and uncertainty
- Hypothesis tests and what a p-value means
- A/B tests, sample size and common biases
- Correlation, causation and confounding
- Supervised Learning, Regression and Classification - Frame a problem for machine learning, split data correctly, build regression and classification models with Scikit-learn, and always compare them against a simple baseline.
- Frame the problem and split train, validation and test data
- Linear regression in Scikit-learn
- Logistic regression for classification
- Decision trees and random forests
- Gradient boosting
- Baselines and model comparison
- Feature Engineering, Pipelines and Data Leakage - Turn raw columns into useful features, build Scikit-learn pipelines that cannot leak information, handle imbalanced data and validate models correctly when time order matters.
- Encoding categories and scaling numbers
- Create and select features
- Handle imbalanced data
- Build Scikit-learn pipelines
- Detect data leakage
- Validate time-ordered data
- Model Evaluation, Tuning and Interpretability - Choose the right metrics, use cross-validation, tune hyperparameters without fooling yourself, study where a model fails and explain its predictions in a clear model report.
- Choose metrics, thresholds and costs
- Measure regression errors
- Cross-validation done correctly
- Hyperparameter tuning without overfitting
- Error analysis and where models fail
- Interpretability and writing a model report
- Unsupervised Learning, Clustering and Dimensionality Reduction - Find structure in unlabelled data with clustering, reduce dimensions with PCA, detect unusual records and judge whether the groups you find are useful or just noise.
- Group customers with k-means
- Hierarchical clustering and DBSCAN
- Choose k and judge the groups
- PCA: see three columns on two axes
- Find unusual rows and review them
- Turn clusters into a segment plan
- Neural Networks and Deep Learning with PyTorch - Learn how neural networks work, write a PyTorch training loop, train models on tabular data and small images, control overfitting and decide when a neural network is worth using over a classical model.
- How neural networks learn
- PyTorch tensors and the training loop
- Neural networks for tabular data
- Overfitting and regularization
- CNN and image basics
- Neural network or classical model
- NLP and Text Data - Clean and tokenize text, build bag-of-words and TF-IDF models, classify text, use embeddings and pretrained transformer models, and judge honestly where text models fail.
- Text cleaning and tokenization
- Bag of words and TF-IDF
- Text classification with a pipeline
- Embeddings and similarity
- Pretrained transformers
- NLP evaluation and limits
- Deployment, MLOps Basics and Responsible AI - Save and load models, serve predictions through a FastAPI service, test and reproduce your work, track experiments, watch for drift and check your models for bias and privacy problems.
- Save and load models safely
- Serve predictions with FastAPI
- Test and reproduce your results
- Track experiments and version your work
- Monitor a live model for drift
- Responsible AI: privacy and bias
- Integrated Capstone, Portfolio and Interview Preparation - Combine everything in one end-to-end project from raw data to a served model and a final report, then prepare your portfolio, resume points and answers to data science interview questions.
- Capstone: from brief to plan
- Capstone: data checks and EDA
- Capstone: modeling and evaluation
- Capstone: API, README and report
- GitHub portfolio and resume points
- Data science interview practice
Frequently asked questions
How long is the Data Science & Machine Learning course?
It runs for 32 weeks, with 12 modules. It is self-paced online learning from EasySkill Career Academy, Surat.
How much does the Data Science & Machine Learning course cost?
The fee is Rs 24,999, or about Rs 4,166 a month on EMI.
What will I build in Data Science & Machine Learning?
Every module ends in a graded practical task that AI checks in about 30 seconds, so you finish with a portfolio you can show to employers or clients.
Which roles does Data Science & Machine Learning prepare me for?
It prepares you to apply for roles such as Junior Data Scientist, Associate Data Scientist, Data Scientist, Machine Learning Engineer, Junior Machine Learning Engineer. It does not guarantee a job or income.
Can I learn online, or do I have to visit a centre?
You can learn fully online. EasySkill also has centres in Adajan, Vesu and Bhestan, Surat, if you prefer to visit.
All career tracks