CLASS 10 AI
CHAPTER - ADVANCED CONCEPTS OF MODELLING IN AI
INTRODUCTION :-
Modeling in AI means creating a system that can learn from data and make
predictions/decisions.
This chapter explores advanced AI modeling concepts like types of learning, evaluation
methods, and bias-ethics considerations.
TYPES OF LEARNING IN AI :-
AI models learn in different ways depending on data and goals. It Consists of Three Types ;
• Supervised Learning
• Unsupervised Learning
• Reinforcement Learning
SUPERVISED LEARNING :-
• Training data has input + correct output (labels).
• The model learns to map inputs → outputs.
Examples :-
i) Predicting house prices (regression).
ii) Email spam detection (classification).
UNSUPERVISED LEARNING :-
• Training data has inputs only (no labels).
• The model finds patterns and groups in data.
Examples :-
i) Customer segmentation.
ii) Grouping news articles by topic.
REINFORCEMENT LEARNING :-
• Model learns through trial and error using rewards & penalties.
Examples :-
i) Training robots.
ii) Game-playing AI (like chess, self-driving cars).
MODEL EVALUATION :-
• To check whether a model works well:
• Training Data → Used to teach the model.
• Testing Data → Used to check accuracy.
Metrics used :-
• Accuracy – Correct predictions/Total predictions.
• Precision & Recall – Quality of predictions.
• Confusion Matrix – Table showing correct vs. incorrect predictions.
OVERFITTING AND UNDERFITTING :-
• Overfitting – Model learns too much from training data (remembers instead of understanding).
Works poorly on new data.
• Underfitting – Model is too simple, cannot learn enough from training data.
• Solution: Use proper data split, cross-validation, and tuning.
BIAS IN AI MODELS :-
• Bias means unfair or inaccurate outcomes due to wrong/imbalanced data.
Causes :-
i) Historical bias in data.
ii) Limited or poor-quality dataset.
Example: AI face recognition working better on certain skin tones.
Solution: Use diverse, balanced, and ethical datasets.
ETHICS IN AI MODELING :-
• AI must be transparent, fair, and accountable.
• Avoid misuse of personal data.
• Follow laws and policies for responsible AI.
IMPORTANCE OF ADVANCED MODELING CONCEPTS :-
• Helps in building reliable, accurate, and fair AI models.
• Ensures AI can be trusted for real-world decision-making.
• Encourages responsible use of AI technology.
CONCLUSION :-
The Advanced Concepts of Modeling focus on how AI learns (supervised, unsupervised,
reinforcement), how we evaluate it, and how to avoid bias, overfitting, underfitting, while
ensuring ethical AI use. This ensures AI models are accurate, fair, and trustworthy.
