End-to-end predictive model
Deliverable: A complete project from raw data to evaluated model, with documented feature engineering, a baseline comparison, cross-validation results and an explicit analysis of where the model fails.
Course
Data Science
Statistics, Python, SQL and machine learning taught as one connected discipline, with the emphasis on framing a problem correctly and knowing when a result is not real.
- Duration: 7 months
- Classroom · Online live
- Level: intermediate
What you will be able to do
- Turn a vague business question into a specific, testable data question
- Build reproducible data pipelines from raw sources to analysis-ready datasets
- Apply and interpret statistical tests without overstating what they prove
- Train, tune and honestly evaluate supervised models on real data
- Detect data leakage, overfitting and spurious correlation in your own work
- Communicate a modelling result, its uncertainty and its limits to a non-technical audience
Who this course is for
- Graduates in engineering, science, mathematics, statistics, economics or commerce
- Working analysts moving from reporting into modelling
- Comfort with school-level mathematics; statistics is taught from first principles
- Some prior programming exposure helps but Python is covered from the start
Syllabus
9 modules · 7 months
Module 1. Statistics that hold up
- Descriptive statistics and distribution shape
- Sampling, sampling error and confidence intervals
- Hypothesis testing and what a p-value does not tell you
- Correlation, causation and confounding variables
- Experiment design and reading an A/B test honestly
Module 2. Python for data work
- Python fundamentals oriented toward data
- pandas indexing, joining, grouping and reshaping
- Handling missing data without quietly biasing the result
- Vectorised thinking with NumPy
- Reproducible notebooks and project structure
Module 3. SQL and data acquisition
- Relational modelling and joins at analytical scale
- Window functions for cohort and time-series work
- Query performance on large tables
- APIs, files and combining heterogeneous sources
Module 4. Exploratory analysis
- Profiling a dataset before modelling it
- Outlier detection and deciding what to do about them
- Feature distributions and transformations
- Visual analysis that reveals rather than decorates
Module 5. Supervised machine learning
- Regression and classification framing
- Linear and logistic models, and reading their coefficients
- Trees, random forests and gradient boosting
- Train, validation and test discipline, and how leakage happens
- Cross-validation and hyperparameter search
- Metric selection and the cost of each error type
Module 6. Unsupervised methods
- Clustering with k-means and hierarchical methods
- Dimensionality reduction with PCA
- Judging whether a cluster is real or an artefact
Module 7. Time series
- Trend, seasonality and stationarity
- Baseline forecasts and why they are hard to beat
- ARIMA-family models
- Backtesting a forecast honestly
Module 8. Communication and deployment basics
- Writing up a model for a decision-maker
- Quantifying and communicating uncertainty
- Model monitoring and drift
- Packaging a model behind a simple interface
Module 9. Capstone and interviews
- Scoping a capstone that can be finished and defended
- Case-study interview practice
- Statistics and modelling question drills
Tools and technologies you will use
- Python
- pandas and NumPy
- scikit-learn
- statsmodels
- SQL
- matplotlib and seaborn
- Jupyter
- Git
Projects you will build
Statistical investigation
Deliverable: An investigation of a real question using hypothesis testing, with stated assumptions, the test chosen and why, effect size, and an honest statement of what the result does not establish.
Customer segmentation
Deliverable: A clustering analysis with justification for the number of clusters, profiling of each segment, and a written argument for whether the segments are actionable or merely statistically distinct.
Forecasting capstone
Deliverable: A time-series forecast benchmarked against a naive baseline, backtested across multiple windows, with the error distribution and the conditions under which the forecast should not be trusted.
Where this course can take you
- Data Scientist
- Machine Learning Engineer
- Quantitative Analyst
- Senior Data Analyst
- Research Analyst
Duration, modes and fees
- Duration
- 7 months
- Delivery modes
- Classroom · Online live
- Fees
- Share your details for the current fee
- Fees vary by batch and delivery mode. Share your details and an advisor will confirm the current fee.
Placement assistance
Every student gets placement assistance — that is what 100% placement assistance means. It is support for all, not a job for all. We do not promise a specific salary, a specific number of interviews, or placement at any named company, and you should be wary of anyone who does.
What is included
- A place in the monthly placement drive, held every third Saturday
- The readiness programme every second Saturday — mock interviews and preparation
- CV review against the specific roles you are targeting
- Portfolio review, so your project work is presented the way a reviewer will read it
- Access to the vacancy pool employers send directly to the Samyak network
- Guidance on which roles realistically fit your background and which do not
- A place in the next drive, with coaching, if you are not selected in this one
What is not included
- Any guarantee of a job, an interview, or a particular salary
- Placement at a named or partner company
- Applying to jobs on your behalf
- Support before you have completed the course and its project work
- Visa, relocation or overseas placement assistance
The skill this course is actually about
It is not model training. Fitting a gradient boosting model is four lines of code and any tutorial will show you.
The skill is knowing whether the number that comes out means anything — whether the training set leaked information from the future, whether the test split respected time order, whether the metric you chose rewards the behaviour you want, whether the effect survives a different sample. Most models that fail in production were technically well-fitted and conceptually wrong.
So statistics comes first here, before any machine learning. Understanding sampling variation and confounding is what lets you look at a 94% accuracy score and ask the right follow-up question.
Baselines, everywhere
Every modelling project in this course requires a baseline first — predict the mean, predict last month, predict the majority class. You are not allowed to report a model result without it.
This is unglamorous and it is the fastest way to develop judgement. A sophisticated model that barely beats “predict last month” is not a success, and noticing that early is what separates useful data science from expensive theatre.
Honest expectations
Seven months gets you genuinely competent at framing problems, building pipelines, modelling carefully and communicating results. It does not make you a research scientist, and it does not substitute for domain knowledge in a specific industry. Many of our learners enter as analysts and move into data science roles from inside a company, which is a well-trodden and realistic path.
Questions
Data Science — frequently asked questions
What is the real difference between data analytics and data science?
Analytics explains what happened and why, using SQL, dashboards and statistics. Data science extends that into prediction and inference — modelling what is likely to happen and quantifying confidence. Analytics roles are more numerous at entry level, so if you are starting from zero we usually suggest analytics first and this course second.
Do I need a mathematics or statistics degree?
No. You need school-level mathematics and willingness to work through the statistics module properly. We teach statistics as applied reasoning rather than proof. What genuinely matters more is scepticism — the instinct to ask whether a result could be an artefact — and that is taught by practice, not by prior credentials.
Will this course get me a data scientist job directly?
It gives you the skills and a defensible portfolio. Whether it gets you the title depends on the market and your background; many people enter as an analyst and move across within a year or two. Anyone promising a data scientist role as an outcome of a course is overselling, and we would rather set the expectation accurately.
How much overlap is there with your AI course?
Some, in the machine learning foundations. This course goes deeper into statistics, inference, experimental design and time series. The AI course goes deeper into deep learning, transformers and generative systems. Choose by the work you want — reasoning from data, or building AI applications.
Is seven months realistic alongside a job?
Yes, and most of our learners do it that way, with weekday evening or weekend batches. Budget around ten hours a week outside class. The statistics and machine learning modules in particular reward practice between sessions rather than passive attendance.
Enquire about Data Science
Three details is all we need. A course advisor will call you back.
Related courses
- Data Analytics
A structured path from spreadsheets to SQL, Power BI and Python, built around the four things a data analyst is actually paid to do — pull the data, clean it, analyse it and explain it.
- Artificial Intelligence
Python, machine learning foundations and applied generative AI in one track — ending with a retrieval-augmented application you have built, evaluated and can explain end to end.
- Python Programming
Python taught as a working tool rather than a syntax tour — you finish able to read unfamiliar code, automate real tasks, call APIs and write tests that catch your own mistakes.
Next step
Talk to a course advisor
Tell us what you want to learn and we will help you pick the right course, batch and mode.