Skip to content

SOTA IOAI Guide INAIO Guide Checklist Tests

Discord

Level

Follows the IOAI 2026 Syllabus

1Getting Started

1.1What is IOAI?

1.1.1 Introduction to the International Olympiad in Artificial Intelligence
1.1.2 IOAI Syllabus and Problem Areas

1.2How to Use This Guide

1.2.1 Guide Structure and Learning Path

1.3Setting Up

1.3.1 Setting Up Python

1.4What is INAIO?

1.4.1 Introduction to the Indian National AI Olympiad

1.5How to Contribute

1.5.1 Contributing to the Guide

2Foundational Skills

2.1Python Basics

2.1.1 Python Fundamentals B
2.1.2 Object-Oriented Programming, Classes, and Inheritance B
2.1.3 Simple Python Problems B

2.2Jupyter Notebook

2.2.1 Jupyter Notebook Interface and Basics B

2.3Mathematics Basics

2.3.1 Terms and Symbols Used in AI B
2.3.2 Probability and Statistics B
2.3.3 Calculus S
2.3.4 Linear Algebra S

2.4Computational Practice

2.4.1 Training on CPU and GPU S
2.4.2 Working with Memory, Batches, and Checkpoints S

3Data Manipulation

3.1Data Manipulation Libraries

3.1.1 Pandas B
3.1.2 Matplotlib B
3.1.3 Seaborn for Visualization S
3.1.4 Data Cleaning B

3.2Preparing Data for Models

3.2.1 Feature Engineering S
3.2.2 Feature Selection S

3.3Data Beyond Tables

3.3.1 Working with Time Series and Variable-Length Data S
3.3.2 Preparing Multimodal Inputs S

4Classical Machine Learning

4.1General Problem Setup

4.1.1 Terminology Used in Classical Machine Learning B
4.1.2 Structure of a Typical Classical Machine Learning Problem B
4.1.3 Approaching a Typical Classical Machine Learning Problem B
4.1.4 From Task Statement to Baseline S
4.1.5 Choosing a Validation Split S
4.1.6 Building a Reliable Tabular Pipeline S
4.1.7 Practical Summary of Classical Machine Learning S

4.2Scikit-learn

4.2.1 Scikit-learn Basics B
4.2.2 Pipelines and Column Transformers S

4.3Classical Machine Learning Theory

4.3.1 Loss Functions B
4.3.2 Underfitting and Overfitting B
4.3.3 Semi-Supervised Learning G

4.4EDA

4.4.1 Useful Graphs to See S
4.4.2 Useful Numbers to See S

4.5Classical Machine Learning Models

4.5.1 Linear Regression B
4.5.2 Logistic Regression B
4.5.3 L1 and L2 Regularization S
4.5.4 Support Vector Machines S
4.5.5 K-Nearest Neighbors B
4.5.6 Decision Trees B
4.5.7 K-Means Clustering B
4.5.8 Principal Component Analysis S
4.5.9 Density-Based and Hierarchical Clustering S
4.5.10 Spectral Clustering G
4.5.11 Ranking Problems and Ranking Models G

4.6Model Ensembles

4.6.1 Introduction to Model Ensembles B
4.6.2 Random Forests B
4.6.3 XGBoost B
4.6.4 LightGBM S
4.6.5 CatBoost S
4.6.6 Choosing a Tabular Model S
4.6.7 Stacking and Blending G

4.7Post-Training Theory

4.7.1 Model Evaluation Metrics B
4.7.2 Cross-Validation B
4.7.3 Hyperparameter Tuning B
4.7.4 Advanced Model Evaluation Metrics and Methods S
4.7.5 Matching Metrics to Tasks S
4.7.6 Error Analysis and Data Leakage S
4.7.7 Domain Shift and Robust Validation G
4.7.8 Uncertainty in Model Comparisons G

4.8Unsupervised Learning and Representations

4.8.1 t-SNE for Exploring Embeddings S
4.8.2 UMAP for Exploring Embeddings G
4.8.3 Clustering and Representation Diagnostics G

4.9Problem Walkthroughs

4.9.1 Titanic Walkthrough B
4.9.2 Feature Engineering Walkthrough S
4.9.3 Feature Selection Walkthrough S
4.9.4 End-to-End Tabular Contest Walkthrough S
4.9.5 Self-Supervision Problem Walkthrough G
4.9.6 Stacking and Ensembling Problem Walkthrough G
4.9.7 Domain Shift Problem Walkthrough G

5Deep Learning

5.1Tensor, Array, and Matrix Manipulation Libraries

5.1.1 NumPy Basics B
5.1.2 PyTorch Basics B

5.2Deep Learning Fundamentals

5.2.1 Multi-Layer Perceptrons B
5.2.2 Gradient Descent B
5.2.3 Backpropagation B
5.2.4 Tensors, Matrices, and Related Mathematical Functions B
5.2.5 Activation Functions B
5.2.6 Loss Functions B
5.2.7 Choosing a Loss Function G
5.2.8 Advanced Loss Functions G
5.2.9 Optimizers: SGD, Momentum, Adam, and AdamW S
5.2.10 Learning Rates and Schedules S
5.2.11 Regularization: Dropout, Early Stopping, and Weight Decay S
5.2.12 Weight Initialization S
5.2.13 Batch Normalization and Layer Normalization S
5.2.14 Diagnosing Training Curves S
5.2.15 Choosing a Model Architecture G
5.2.16 Autoencoders G
5.2.17 Contrastive Learning G
5.2.18 Knowledge Distillation G

5.3PyTorch Deep Learning Pipeline

5.3.1 Dataset and DataLoader B
5.3.2 A Simple MLP in PyTorch B
5.3.3 PyTorch Pipeline for a Simple Machine Learning Problem B
5.3.4 Simple Problem Examples B
5.3.5 A Reusable Training and Evaluation Loop S
5.3.6 Saving, Loading, and Resuming Models S
5.3.7 Using a Pretrained Model S
5.3.8 Full and Parameter-Efficient Fine-Tuning G
5.3.9 LoRA and PEFT G
5.3.10 GPU and Memory Optimization G

5.4Sequence Models

5.4.1 Recurrent Neural Networks S
5.4.2 Long Short-Term Memory Networks S

5.5Representations and Attention

5.5.1 Embeddings Across Text, Image, and Audio S
5.5.2 Using Pretrained Models S
5.5.3 Attention Mechanisms G
5.5.4 Transformers G
5.5.5 Self-Supervised and Contrastive Pretraining G

6Computer Vision

6.1Image Modality Basics

6.1.1 Basic Image Terminology B
6.1.2 Structure and Types of Computer Vision Problems S

6.2Computer Vision Theory

6.2.1 Convolutional Layers B
6.2.2 Pooling B

6.3Raw Image Processing

6.3.1 Reading and Manipulating Images with Pillow B
6.3.2 Loading Images into Tensors with Torchvision B
6.3.3 Preparing Images, Masks, and Boxes S
6.3.4 Image Augmentation S

6.4Computer Vision Models

6.4.1 Convolutional Neural Networks B
6.4.2 MNIST B
6.4.3 Image Classification with Pretrained Encoders S
6.4.4 Evolution of Computer Vision Architectures G
6.4.5 ResNet S
6.4.6 EfficientNet S
6.4.7 ConvNeXt S
6.4.8 U-Net S
6.4.9 YOLO S
6.4.10 Comparing Detection Models: SSD and DETR G
6.4.11 CLIP and Image-Text Retrieval S
6.4.12 Vision Transformers G
6.4.13 Self-Supervised Vision G
6.4.14 Generating Images with GANs G
6.4.15 Diffusion Models G

6.5Vision Problem Solving

6.5.1 A Vision Task Baseline and Submission S
6.5.2 The Vision Contest Pipeline G
6.5.3 Segmentation or Detection Walkthrough G
6.5.4 Vision Contest Problem Walkthrough G

7Natural Language Processing

7.1Text Modality

7.1.1 Text Modality Basics B
7.1.2 Structure and Types of NLP Problems S

7.2Text Processing

7.2.1 Text Preprocessing B
7.2.2 Tokenization B
7.2.3 Padding, Truncation, and Attention Masks S

7.3Basic NLP Algorithms

7.3.1 Bag of Words B
7.3.2 N-grams B
7.3.3 TF-IDF B
7.3.4 Vector Embeddings of Simple Text Features B
7.3.5 Word2Vec and GloVe S
7.3.6 Using Text Encoders for Classification and Retrieval S
7.3.7 BERT S

7.4Language Models and Text Tasks

7.4.1 Text Classification with a Pretrained Model S
7.4.2 Using a Causal Language Model S
7.4.3 Encoder-Decoder Models in Practice S
7.4.4 Natural Language Inference S
7.4.5 How Causal Transformers Work G
7.4.6 Retrieval-Augmented Generation G
7.4.7 Choosing Between Encoder, Encoder-Decoder, and Causal Models G

7.5NLP Problem Walkthroughs

7.5.1 Text Classification Walkthrough S
7.5.2 Retrieval or NLI Walkthrough G

8Audio

8.1Audio Modality Basics

8.1.1 Waveforms, Sampling Rate, Duration, Amplitude, and Channels B
8.1.2 Theory of Audio Features B
8.1.3 Structure and Types of Audio Problems S

8.2Librosa

8.2.1 Loading Audio with Librosa B
8.2.2 Audio Feature Extraction with Librosa B

8.3Audio Models

8.3.1 Audio Classification and a Baseline S
8.3.2 HuBERT S
8.3.3 Audio Spectrogram Transformers G
8.3.4 Using Whisper for Transcription S
8.3.5 Comparing Whisper, Qwen-Audio, and Voxtral G
8.3.6 Speaker Diarization G
8.3.7 Audio Contest Problem Walkthrough G

9Multimodal Learning

9.1Vision-Language Models

9.1.1 Using BLIP for Captioning and Retrieval S
9.1.2 Vision-Language Problem Solving G
9.1.3 Multimodal Problem Walkthrough G

9.2Audio-Text Models

9.2.1 CLAP S

10Contest Practice and Strategy

10.1General Contest Modules

10.1.1 General Contest Module S

0
1
2
3
4
5
6
7
8
9
10

Your path to IOAI

Start here

Scroll to move · pinch or ctrl + scroll to zoom · click an area to open it

On this page

© 2026 SOTA – AI Community. A student-led community, run by volunteers.

SOTA IOAI Guide Checklist Tests Discord LinkedIn Email Privacy Terms