Assam HS Skill Subject Artificial Intelligence study plan

Define artificial intelligence, machine learning and deep learning in your own words and draw a diagram showing how the three sets nest inside one another

Trace five milestones in AI history from the 1950s to the present and write one sentence on why each mattered

Set up a Python environment with NumPy and pandas installed and run a script that prints their versions

Write and run Python code using integers, floats, strings, lists and dictionaries, then convert between at least three of these types

Classify five everyday products (search engine, spam filter, chess engine, face unlock, recommendation feed) as rule-based AI, ML or DL and justify each label

Write loops and conditional statements that filter and transform a list of at least 20 numeric values

Define three functions with parameters and return values, including one that computes a mean without using a library

Load a CSV file into a pandas DataFrame and inspect it with head, shape, dtypes and describe

Perform column selection, filtering with boolean masks, and grouping with groupby on the loaded dataset

Create a NumPy array, reshape it, and compute element-wise and aggregate statistics

Distinguish structured, semi-structured and unstructured data and give two examples of each from Indian public data sources

Identify and handle missing values, duplicates and inconsistent text entries in a messy dataset

Compute and interpret mean, median, mode, range and standard deviation for two numeric columns

Plot a histogram, a bar chart and a scatter plot with labelled axes and explain what each reveals

Write a short note on data collection ethics covering consent and anonymisation

Explain the difference between supervised and unsupervised learning and place five tasks into each category

Train a classification model on a small labelled dataset and report its accuracy

Train a linear regression model and interpret the slope and intercept in plain language

Build a confusion matrix by hand for 20 predictions and compute precision, recall and F1 score

Describe why a train-test split is needed and demonstrate the effect of testing on training data

Draw a single perceptron and label inputs, weights, bias and activation output

Compute the output of a perceptron by hand for three input vectors using a step function

Compare sigmoid, ReLU and softmax activation functions and state where each is typically used

Describe in writing how forward propagation and loss-driven weight updates work, without deriving formulas

Explain overfitting and list two techniques that reduce it

Preprocess text by lowercasing, removing punctuation, tokenising and removing stopwords from a set of 20 sentences

Convert a small text corpus into bag-of-words and TF-IDF representations and explain the difference

Design a rule-based chatbot with at least eight intents and test it against ten sample inputs

Describe the steps of an image classification pipeline from pixels to predicted label

List three real NLP applications and three real computer vision applications used in India today

Identify three sources of bias in a dataset or model and propose a mitigation for each

Explain the purpose of consent, purpose limitation and data minimisation under Indian data protection law in your own words

Analyse a case where an AI system caused harm and state which ethical principle was violated

Argue both for and against the claim that AI will displace more jobs than it creates in the next decade

Draft a five-point responsible AI checklist for a school AI project

Frame a capstone problem with a clear input, output, success metric and stated limitations

Prepare and document a dataset, recording its source, cleaning steps and known biases

Build and evaluate a baseline model, then record its results against the success metric

Write project documentation covering problem, method, results and limitations in under four pages

Rehearse and deliver a five-minute presentation with slides and answer three likely questions