Python, SQL, statistics, data analysis, machine learning, evaluation and a reproducible capstone.
What you will cover
Python
SQL
statistics
data analysis
machine learning
evaluation
a reproducible capstone
Topic list from the institute's programme record.
Certification
Certificate terms for this programme are confirmed by the institute at admission. Any ADIVAS course-completion certificate is an institute certificate; it is not a university degree, an externally accredited qualification or a vendor certification (such as AWS or Microsoft) unless the institute states so in writing.
Curriculum 0.1 · Draft: written for ADIVAS and awaiting academic review and approval. Topics and weeks may change before your batch starts.
Your skill trail
You move through 7 stations. Each week builds one skill you can show, not just a topic you have heard.
1
Python foundations
Write, read and debug Python programs for data tasks.
Week 1 Python basics: values, variables and control flow
Week 2 Functions, lists and dictionaries
Week 3 Files, text processing and testing
Week 4 NumPy, environments and Git
2
SQL and data access
Retrieve and combine data from relational databases.
Week 5 SQL fundamentals
Week 6 Joins and SQL from Python
3
Data analysis with pandas
Explore, clean, transform and visualise data to answer questions.
Week 7 pandas: loading and exploring data
Week 8 Cleaning and transforming data
Week 9 Visualisation and communication
Week 10 Exploratory data analysis mini-project
4
Statistics for decisions
Use probability and inference to make careful claims from data.
Week 11 Descriptive statistics and probability
Week 12 Distributions and sampling
Week 13 Confidence intervals and hypothesis tests
Week 14 Correlation, regression and linear models
5
Machine learning
Train and tune simple models and avoid common mistakes.
Week 15 Machine learning workflow
Week 16 Regression models
Week 17 Classification models
Week 18 Ensembles and tuning
Week 19 Unsupervised learning and text basics
6
Evaluation, ethics and communication
Judge models honestly and report results responsibly.
Week 20 Model evaluation in depth
Week 21 Fairness, privacy and misuse
Week 22 Reporting and presenting results
7
Reproducible capstone
Plan, build and present a reproducible end-to-end project.
Week 23 Capstone build
Week 24 Capstone evaluation and presentation
★
Predicting Student Course Completion
Your project, reviewed by a trainer.
The ADIVAS mastery ladder: how each skill is recognised
Explored You have seen the idea and tried a first example.
Practising You can do it with hints and keep improving.
Proficient You complete the weekly practical task correctly on your own.
Proven You apply it in a project reviewed by your trainer.
Skill-level tracking is being added to the learner workspace. Your trainer already reviews weekly practical work.
What you will be able to do
Write Python programs with functions, collections and files, and read, debug and test short scripts.
Query and combine relational data with SQL and load it into Python for analysis.
Explore and clean data with pandas, producing summary tables and charts that answer a stated question.
Apply descriptive statistics, probability, sampling, confidence intervals and hypothesis tests, and interpret them correctly.
Build regression and classification models with scikit-learn using proper train, validation and test splits.
Evaluate models with suitable metrics, spot overfitting and leakage, and compare against a simple baseline.
Examine fairness, privacy and misuse risks in data projects and document limits and assumptions.
Deliver a reproducible capstone with version control, a fixed environment, a README and a clear report.
Who it suits
School Students in Classes 11-12 with an interest in mathematics or computer science who want to see how data, patterns and predictions work in plain terms.
College Science, engineering and mathematics students who want to build real analysis and modelling skills for projects and further study.
Professional Analysts, engineers and career changers who want to move from spreadsheets to Python-based analysis and applied machine learning.
Prerequisites: School-level algebra and percentages. Comfort with a computer. Programming is taught from the start, but regular practice is expected. Tools: Python 3 with a virtual environment, Jupyter Notebook or JupyterLab and VS Code, pandas, NumPy, Matplotlib, SciPy and scikit-learn, SQLite or PostgreSQL, Git and a hosted repository, Provided synthetic or public-domain datasets.
Weekly plan
Python foundations Weeks 1–4
Week 1: Python basics: values, variables and control flow
Theory
Numbers, strings, booleans and type conversion
Variables and expressions
if, elif, else and comparison operators
for and while loops
Lab
Write a script that reads marks typed by the user, computes total and percentage, and prints a grade; test it with five inputs including a boundary case.
Homework
Write a FizzBuzz-style program for a local context, such as labelling bill amounts.
Deliverable
Two small scripts with a table of test inputs and expected outputs.
Weekly checkpoint
Trace a short loop by hand; fix an off-by-one error.
You will be able to
Write Python programs with functions, collections and files, and read, debug and test short scripts.
Week 2: Functions, lists and dictionaries
Theory
Defining functions, parameters and return values
Lists, tuples, sets and dictionaries
List comprehensions
Reading error messages and tracebacks
Lab
Write functions to compute mean, median and mode of a list of exam marks, and test each with an even count, an odd count and an empty list.
Homework
Store student marks in a dictionary and write a function that returns the top three names.
Deliverable
A small statistics module with a test table.
Weekly checkpoint
Predict output of dictionary and slicing code; debug a failing function.
You will be able to
Write Python programs with functions, collections and files, and read, debug and test short scripts.
Week 3: Files, text processing and testing
Theory
Reading and writing text and CSV files
String methods and simple regular expressions
Exceptions: try and except
assert statements and tiny unit tests
Lab
Read a CSV of attendance with some bad rows, skip and report them, and write a clean file; add three assert tests.
Homework
Count word frequencies in a Tamil or English paragraph and print the top five.
Deliverable
A script that cleans a CSV, with a log of rejected rows.
Weekly checkpoint
Find the bug in a file-reading snippet; write a test for a function.
You will be able to
Write Python programs with functions, collections and files, and read, debug and test short scripts.
Week 4: NumPy, environments and Git
Theory
Arrays, vectorised operations and broadcasting
Virtual environments and requirements files
Git: commit, branch, merge and history
Notebooks versus scripts
Lab
Compute z-scores of exam marks with NumPy without loops, initialise a Git repository with a requirements file, and make three meaningful commits.
Homework
Recreate the environment on another machine or folder from requirements.txt.
Deliverable
A repository with a README that lets another person run the code.
Weekly checkpoint
Explain why vectorisation is faster; read a git log and describe the change history.
You will be able to
Write Python programs with functions, collections and files, and read, debug and test short scripts. Deliver a reproducible capstone with version control, a fixed environment, a README and a clear report.
SQL and data access Weeks 5–6
Week 5: SQL fundamentals
Theory
Tables, keys and relationships
SELECT, WHERE, ORDER BY, LIMIT
Aggregation with GROUP BY and HAVING
Handling NULL
Lab
Load a retail SQLite database and answer ten business questions with SQL, checking row counts.
Homework
Translate six plain-English questions into queries.
Deliverable
A tested SQL file with a note of the expected result for each query.
Weekly checkpoint
Read four queries and say what they return.
You will be able to
Query and combine relational data with SQL and load it into Python for analysis.
Week 6: Joins and SQL from Python
Theory
INNER and LEFT JOIN
CTEs and subqueries
Window functions: ranking and running totals
Parameterised queries and loading results into pandas
Lab
Join orders, customers and products; compute monthly revenue and rank products within category; load the result into a DataFrame using a parameterised query.
Homework
Explain why string-built queries are unsafe and show the parameterised version.
Deliverable
A notebook mixing SQL and pandas with checks that totals agree.
Weekly checkpoint
Predict row counts of a join; spot an unsafe query.
You will be able to
Query and combine relational data with SQL and load it into Python for analysis. Explore and clean data with pandas, producing summary tables and charts that answer a stated question.
Data analysis with pandas Weeks 7–10
Week 7: pandas: loading and exploring data
Theory
Series and DataFrame
Selecting, filtering and sorting
Summary statistics and value counts
Reading CSV, Excel and SQL
Lab
Explore a public-style dataset of school results: shape, types, missing counts and five questions answered with code.
Homework
Write a one-page data dictionary for the dataset.
Deliverable
An exploration notebook with questions, code and conclusions.
Weekly checkpoint
Read pandas snippets and state results.
You will be able to
Explore and clean data with pandas, producing summary tables and charts that answer a stated question.
Week 8: Cleaning and transforming data
Theory
Missing values: drop, fill, flag
Duplicates, type conversion and text standardisation
Dates and time zones
groupby, merge, pivot and melt
Lab
Clean a messy customer-orders extract, document each decision in a log, and produce a tidy table.
Homework
Re-run the cleaning on a fresh copy of the raw file to prove it is repeatable.
Deliverable
A cleaning script plus a data-quality log.
Weekly checkpoint
Choose and justify a method for missing values in three cases.
You will be able to
Explore and clean data with pandas, producing summary tables and charts that answer a stated question.
Week 9: Visualisation and communication
Theory
Matplotlib structure: figure, axes, labels
Histogram, box plot, scatter, bar and line
Choosing charts and avoiding misleading scales
Annotating findings
Lab
Create five labelled charts that answer separate questions, then redraw one misleading chart honestly.
Homework
Write a caption for each chart stating what it shows and what it does not show.
Deliverable
A figure set with captions and a short note on design choices.
Weekly checkpoint
Interpret a box plot and a scatter plot; critique a poor chart.
You will be able to
Explore and clean data with pandas, producing summary tables and charts that answer a stated question. Apply descriptive statistics, probability, sampling, confidence intervals and hypothesis tests, and interpret them correctly.
Week 10: Exploratory data analysis mini-project
Theory
Framing a question and hypotheses
Segments, trends and outliers
Correlation, confounding and Simpson's paradox
Writing an insight note
Lab
Analyse a bike-rental-style dataset for the effect of weather and weekdays on demand; present three findings, each with a caveat.
Homework
Peer review a classmate's notebook using a checklist.
Deliverable
An EDA notebook and a one-page findings note.
Weekly checkpoint
Interpret a table where an overall trend reverses within groups.
You will be able to
Explore and clean data with pandas, producing summary tables and charts that answer a stated question. Apply descriptive statistics, probability, sampling, confidence intervals and hypothesis tests, and interpret them correctly.
Statistics for decisions Weeks 11–14
Week 11: Descriptive statistics and probability
Theory
Mean, median, standard deviation, quartiles
Skew and when the mean misleads
Probability rules and conditional probability
Bayes' rule with a medical-test example
Lab
Simulate dice and coin experiments in Python; compute probabilities by simulation and by formula; work through a screening-test example.
Homework
Explain base-rate error to a friend in three sentences.
Deliverable
A notebook comparing simulated and theoretical probabilities.
Weekly checkpoint
Compute a conditional probability from a two-by-two table.
You will be able to
Apply descriptive statistics, probability, sampling, confidence intervals and hypothesis tests, and interpret them correctly.
Week 12: Distributions and sampling
Theory
Normal, binomial and uniform distributions
Sampling, bias and the central limit theorem
Standard error
Random seeds and reproducibility
Lab
Draw repeated samples from a skewed dataset, plot the distribution of sample means and observe the central limit theorem; show how a biased sample misleads.
Homework
Describe a sampling plan for a survey in your college that avoids obvious bias.
Deliverable
A sampling notebook with plots and a plan.
Weekly checkpoint
Identify the bias in three survey scenarios.
You will be able to
Apply descriptive statistics, probability, sampling, confidence intervals and hypothesis tests, and interpret them correctly.
Week 13: Confidence intervals and hypothesis tests
Theory
Confidence intervals and what they mean
Null and alternative hypotheses, p-values and their limits
t-test, chi-square test and A/B comparisons
Effect size, power and multiple testing
Lab
Analyse an A/B test on two versions of a sign-up page: compute interval and p-value, then explain whether the difference matters in practice.
Homework
Write what a p-value of 0.03 does and does not tell you.
Deliverable
An A/B analysis report with an effect size and a caution.
Weekly checkpoint
Interpret a confidence interval that includes zero.
You will be able to
Apply descriptive statistics, probability, sampling, confidence intervals and hypothesis tests, and interpret them correctly.
Week 14: Correlation, regression and linear models
Theory
Correlation and its limits
Simple and multiple linear regression
Coefficients, residuals and assumptions
Multicollinearity and extrapolation
Lab
Fit a regression of house price on area and age; read coefficients, plot residuals and flag where the model should not be trusted.
Homework
Predict a value inside and outside the data range and discuss which is safer.
Deliverable
A regression notebook with residual analysis.
Weekly checkpoint
Read a coefficient table and state the meaning in words.
You will be able to
Apply descriptive statistics, probability, sampling, confidence intervals and hypothesis tests, and interpret them correctly. Build regression and classification models with scikit-learn using proper train, validation and test splits.
Machine learning Weeks 15–19
Week 15: Machine learning workflow
Theory
Supervised and unsupervised learning
Features, labels and the train, validation and test split
Baselines and why they matter
Data leakage
Lab
Build a baseline predictor (mean or majority class), split data correctly and measure the baseline score; find a deliberately planted leak in a dataset.
Homework
List three leakage examples from real life.
Deliverable
A baseline and split notebook with a leakage note.
Weekly checkpoint
Identify the leak in a given feature list.
You will be able to
Build regression and classification models with scikit-learn using proper train, validation and test splits. Evaluate models with suitable metrics, spot overfitting and leakage, and compare against a simple baseline.
Week 16: Regression models
Theory
Linear regression with scikit-learn
Regularisation: ridge and lasso
Feature scaling and encoding categories
Pipelines
Lab
Predict used-bike prices with a pipeline that scales numbers and encodes brands; compare with the baseline using cross-validation.
Homework
Explain why scaling is fitted on training data only.
Deliverable
A pipeline notebook with comparison table.
Weekly checkpoint
Order the steps of a leak-free pipeline.
You will be able to
Build regression and classification models with scikit-learn using proper train, validation and test splits. Evaluate models with suitable metrics, spot overfitting and leakage, and compare against a simple baseline.
Week 17: Classification models
Theory
Logistic regression
Decision trees and nearest neighbours
Class imbalance
Probability thresholds
Lab
Predict loan default on a synthetic dataset with logistic regression and a tree; try a different threshold and discuss the trade-off.
Homework
Describe a case where a false negative is worse than a false positive.
Deliverable
A classification notebook with confusion matrices.
Weekly checkpoint
Compute precision and recall from a confusion matrix.
You will be able to
Build regression and classification models with scikit-learn using proper train, validation and test splits. Evaluate models with suitable metrics, spot overfitting and leakage, and compare against a simple baseline.
Week 18: Ensembles and tuning
Theory
Random forests and gradient boosting
Hyperparameters and cross-validation
Grid and random search
Feature importance and its cautions
Lab
Tune a random forest with cross-validation, then check the untouched test set once; record how much the score changed from baseline.
Homework
Explain why tuning on the test set gives a dishonest score.
Deliverable
A tuning log with validation and final test scores.
Weekly checkpoint
Spot test-set misuse in a described workflow.
You will be able to
Build regression and classification models with scikit-learn using proper train, validation and test splits. Evaluate models with suitable metrics, spot overfitting and leakage, and compare against a simple baseline.
Week 19: Unsupervised learning and text basics
Theory
k-means clustering and choosing k
Principal component analysis for visualisation
Bag-of-words and TF-IDF
Limits of clusters: they need human interpretation
Lab
Cluster customers by spend and visits and give each cluster a plain-language description; classify short feedback comments with TF-IDF and logistic regression.
Homework
Write what could go wrong if a business acts on the clusters unchecked.
Deliverable
A clustering notebook and a text classifier with error examples.
Weekly checkpoint
Interpret cluster centres from a table.
You will be able to
Build regression and classification models with scikit-learn using proper train, validation and test splits.
Evaluation, ethics and communication Weeks 20–22
Week 20: Model evaluation in depth
Theory
Accuracy, precision, recall, F1, ROC-AUC
Regression metrics: MAE, RMSE, R-squared
Overfitting and learning curves
Calibration and uncertainty
Lab
Evaluate a model on an imbalanced dataset, show that high accuracy hides poor recall, and choose a metric that matches the business cost.
Homework
Write a metric choice memo for a fraud and a spam use case.
Deliverable
An evaluation notebook and a metric memo.
Weekly checkpoint
Choose a metric for four scenarios and justify.
You will be able to
Evaluate models with suitable metrics, spot overfitting and leakage, and compare against a simple baseline.
Week 21: Fairness, privacy and misuse
Theory
Sources of bias in data and labels
Checking performance across groups
Privacy: anonymisation limits, consent and minimisation
Misuse, over-trust and human oversight
Lab
Compare error rates of a model across two groups in a synthetic hiring dataset; list mitigation options and what you would not automate.
Homework
Write a model card summarising purpose, data, limits and risks.
Deliverable
A group-wise evaluation table and a model card.
Weekly checkpoint
Read a group error table and decide if the model is acceptable to deploy.
You will be able to
Examine fairness, privacy and misuse risks in data projects and document limits and assumptions. Evaluate models with suitable metrics, spot overfitting and leakage, and compare against a simple baseline.
Week 22: Reporting and presenting results
Theory
Structuring a technical report
Explaining models to non-technical readers
Uncertainty and limits stated plainly
Peer review and critique
Lab
Turn an earlier model into a two-page report for a shop owner or school head, with one chart and one decision it supports; exchange for peer review.
Homework
Revise the report using the peer comments.
Deliverable
A revised two-page report.
Weekly checkpoint
Rewrite a jargon-heavy paragraph for a manager.
You will be able to
Examine fairness, privacy and misuse risks in data projects and document limits and assumptions. Deliver a reproducible capstone with version control, a fixed environment, a README and a clear report.
Reproducible capstone Weeks 23–24
Week 23: Capstone build
Theory
Project scoping and risk list
Folder structure, configuration and random seeds
Automated checks and a one-command run
Documenting data sources and licences
Lab
Build your capstone from raw data to evaluated model in a version-controlled repository, with a README and a single command that reproduces results.
Homework
Ask a classmate to run your repository from scratch and report problems.
Deliverable
A repository that reproduces the main result.
Weekly checkpoint
Mid-point code review.
You will be able to
Write Python programs with functions, collections and files, and read, debug and test short scripts. Query and combine relational data with SQL and load it into Python for analysis. Explore and clean data with pandas, producing summary tables and charts that answer a stated question. Build regression and classification models with scikit-learn using proper train, validation and test splits. Deliver a reproducible capstone with version control, a fixed environment, a README and a clear report.
Week 24: Capstone evaluation and presentation
Theory
Final evaluation on the untouched test set
Limitations, ethics and next steps
Presenting in ten minutes with one failure case shown
Lab
Run the final evaluation, complete the model card and report, and present with a live reproduction of one result.
Homework
Publish a short reflection on what you would do differently.
Deliverable
Final repository, report, model card and a ten-minute presentation.
Weekly checkpoint
Capstone viva and rubric scoring.
You will be able to
Apply descriptive statistics, probability, sampling, confidence intervals and hypothesis tests, and interpret them correctly. Evaluate models with suitable metrics, spot overfitting and leakage, and compare against a simple baseline. Examine fairness, privacy and misuse risks in data projects and document limits and assumptions. Deliver a reproducible capstone with version control, a fixed environment, a README and a clear report.
Sample lessons
Week 15
Train, validation, test and the honest baseline
A model that is tested on the same data it learned from will always look good. To judge it fairly, you split the data: the training set to learn, a validation set to tune choices, and a test set kept untouched until the end. You should also build a baseline, such as always predicting the most common class, so that you know what score a model must beat. Many apparent successes are only a baseline in disguise or the result of leakage.
1Define target
2Split train and test
3Baseline score
4Train on train only
5Validate and tune
6Test once
The test set is opened once, at the end.
Step by step
Define the target and what you will predict at prediction time.
Split the data into train and test before any exploring or scaling, with a fixed random seed.
Compute a baseline score on the validation data.
Train a model on the training data only.
Compare the validation score with the baseline and tune choices.
Score the test set exactly once and report it with its limits.
In real life
A loan officer wants a model that flags risky loans. If only 3 percent default, a model that says no loan will default is 97 percent accurate and still useless. The baseline exposes this.
Worked example
Problem. A dataset has 1,000 learners, 950 who completed and 50 who dropped out. A model has 96 percent accuracy. Is it good?
Solution. The majority-class baseline predicts every learner will complete and scores 95 percent. The model's 96 percent is only one point better, and we do not know whether it catches any of the 50 dropouts. Check recall on the dropout class: if the model finds only 5 of 50, recall is 10 percent and the model is not useful for the actual purpose.
Common mistakes
Scaling or selecting features on the full dataset before splitting, which leaks information.
Tuning repeatedly on the test set.
Reporting accuracy on imbalanced data without a baseline or recall.
Try it yourself
Compute the baseline accuracy for a dataset with 88 percent majority class.
Split a dataset with a fixed seed and show that two runs give the same split.
List three features that would leak the target in a dropout prediction problem.
Week 13
Reading a hypothesis test and a confidence interval
A hypothesis test asks: if there were really no difference, how surprising would our data be? The p-value measures that surprise; it is not the probability that the claim is true. A confidence interval gives a range of plausible values for the true effect. A small p-value with a tiny effect can be statistically significant but useless in practice, while a large effect from a very small sample may be noise. Always report the size of the effect and its uncertainty.
Result
Interval for difference
Reasonable conclusion
Small p, large effect
+5 to +9 points
Likely real and meaningful
Small p, tiny effect
+0.1 to +0.4 points
Real but possibly not worth acting on
Large p, wide interval
-3 to +9 points
Inconclusive; collect more data
Large p, narrow interval around 0
-0.5 to +0.5 points
Little or no difference
Read p-values together with effect size.
Step by step
State the question and the null hypothesis, such as no difference between page A and page B.
Choose the test that fits the data type, for example a two-proportion test for conversion rates.
Compute the effect (difference) and the confidence interval.
Compute the p-value and compare with the threshold fixed beforehand.
Judge practical importance: is the difference big enough to matter?
Write the conclusion with the interval and the limits of the data.
In real life
A coaching centre tests two enquiry forms. Form B converts 52 of 400 visitors and form A converts 40 of 400. The difference looks promising but needs to be examined with uncertainty before changing the website.
Worked example
Problem. Form A: 40 of 400 enquire (10 percent). Form B: 52 of 400 (13 percent). The 95 percent interval for the difference is -0.5 to +6.5 percentage points. What do you conclude?
Solution. The interval includes zero, so the data are compatible with no difference or with a gain of up to 6.5 points. We cannot claim B is better yet. Collect more data or accept that the effect, if any, is small, and say so plainly instead of announcing a winner.
Common mistakes
Reading p = 0.03 as a 97 percent chance that the hypothesis is true.
Running many tests and reporting only the one that was significant.
Stopping the test when the result first looks good.
Try it yourself
Explain in two sentences what a 95 percent confidence interval means.
Compute the difference and an approximate interval for two given conversion rates.
Decide whether a significant 0.2 percent difference is worth acting on and explain.
Try the questions
The same skills are asked at three levels: School (Classes 6–12), College and Professional. Switch the level and the language (English or Tamil), and say how sure you are before you check: knowing what you are unsure about is part of learning.
Classes 6–12: everyday examples, no prior knowledge. Practice only. Nothing here is graded or saved.
Question 1 / 4 · Compare mean and median
Five friends have pocket money of Rs 100, 100, 120, 130 and Rs 2,000. Which number describes a typical friend better?
Projects
Predicting Student Course Completion
A training centre has a synthetic table of 2,000 learners with attendance percentage, assignments submitted, weekly study hours, course type and whether they completed. Build a reproducible analysis and a classifier to flag learners at risk of dropping out, and evaluate it honestly including fairness checks.
Tasks
Clean and explore the data and write a data-quality log.
Establish a baseline (majority class) and a leak-free train, validation and test split.
Train logistic regression and a tree-based model with cross-validation and compare with the baseline.
Report precision, recall and a confusion matrix; choose a threshold and justify it with the cost of a missed learner.
Compare error rates across course types and write a model card with limits and recommended human follow-up.
A repository that reruns in one command, a table of metrics against the baseline, a group-wise error table and a one-page model card.
How it is assessed
Criterion
Weight
Data cleaning and leak-free design
20%
Modelling with baseline and validation
25%
Evaluation and metric choice
20%
Fairness, privacy and limits
15%
Reproducibility and communication
20%
Stretch challenge: Add a calibration plot and explain how staff should use the risk score without treating it as a fact.
How you are assessed
Each week: five short questions mixing code reading, data interpretation and one chart or table question, plus a notebook or script checked by the trainer against a rubric.
Practical work. Learners submit notebooks, SQL files, scripts with tests and a final repository. Results are judged on correctness, honest evaluation against baselines, documentation and the ability to reproduce results from a clean environment.
What trainers look for
Weight
Correct, reproducible code and analysis
30%
Statistical reasoning and interpretation
20%
Model evaluation and baselines
20%
Ethics, privacy and limits
15%
Communication and documentation
15%
Where this can lead
Prepares you for junior data analyst and entry-level data science support roles that use Python, SQL and statistics.
Prepares you to build and evaluate simple machine-learning models and explain their limits.
Builds a base for postgraduate study or deeper work in machine learning and statistics.
These describe skills you will build. ADIVAS does not guarantee jobs, placement or salary.
Other programmes in Artificial Intelligence & Data Science
