1. Machine Learning (ML)
AI systems that learn from data to improve performance without being explicitly programmed. Subsets include supervised, unsupervised, and reinforcement learning.
2. Deep Learning
A subset of ML using neural networks with many layers (deep neural networks). It's used in tasks like image recognition, natural language processing, and speech recognition.
3. Neural Networks
Inspired by the human brain, these are the foundation of deep learning. They consist of layers of interconnected nodes (neurons) that process data.
4. Natural Language Processing (NLP)
The field of AI that enables machines to understand, interpret, and generate human language. Examples: chatbots, language translation.
5. Computer Vision
AI that enables machines to interpret and understand visual information from the world (e.g., object detection, facial recognition).
6. Reinforcement Learning
A learning method where an agent learns to make decisions by receiving rewards or penalties for actions taken in an environment.
7. Supervised Learning
Machine learning where the model is trained on labeled data. Used in classification and regression tasks.
8. Unsupervised Learning
Learning from unlabeled data to identify hidden patterns (e.g., clustering, dimensionality reduction).
9. Generative AI
AI models that can create new content—text, images, music, etc. Examples: ChatGPT, DALL·E, Midjourney.
10. Transformers
A deep learning architecture especially effective for NLP tasks. Used in models like GPT, BERT, and T5.
11. Large Language Models (LLMs)
Extremely large neural networks trained on massive datasets to understand and generate human-like text (e.g., GPT-4, Claude, Gemini).
12. Bias in AI
The presence of unfair or prejudiced results in AI due to biased training data or model design. Critical in ethical AI development.
13. Semantic Search
Search based on meaning rather than exact keyword matching. Powered by embeddings and vector databases.
14. Overfitting & Underfitting
Overfitting: Model learns the training data too well, including noise.
Underfitting: Model is too simple to capture the underlying pattern.
15. Transfer Learning
Reusing a pre-trained model on a new task with limited data. Saves time and computing resources.
16. Prompt Engineering
Designing effective inputs ("prompts") to get desired outputs from AI models, especially LLMs like ChatGPT.
17. AI Ethics
Study of the moral implications of AI systems—privacy, fairness, accountability, and impact on jobs and society.
18. Data Annotation & Labeling
The process of tagging data (text, image, video) to make it usable for supervised learning.
19. Edge AI
Running AI models locally on devices (phones, IoT, etc.) instead of the cloud. Important for speed, privacy, and offline use.
20. AI Model Evaluation Metrics
Ways to measure how well an AI model performs:
Accuracy, Precision, Recall, F1-score for classification.
MSE, MAE, R² for regression.
BLEU, ROUGE, Perplexity for NLP.
21. Fine-tuning
Adjusting a pre-trained model on domain-specific data to improve performance in specialized applications.
22. Embeddings
Low-dimensional vector representations of high-dimensional data (text, image, audio). These capture semantics and are essential for tasks like similarity search and recommendations.
23. Vector Search
A method of searching based on similarity between embeddings using techniques like cosine similarity or Euclidean distance. Used in semantic search and retrieval-augmented generation.
24. Retrieval-Augmented Generation (RAG)
Combines vector search with generative models by retrieving relevant documents to enhance generation. Key in enterprise chatbots and LLM applications.
25. GPU/TPU Infrastructure
High-performance compute units essential for training large AI models. TPUs (Google) are specialized hardware for tensor operations.
Top comments (0)