Key Concepts & Self-Assessment15 Key Facts
Review key Artificial Intelligence, Machine Learning & Neural Network Foundations exam facts and rate your mastery to track revision.
Progress: 0/15 Rated 0 Mastered 0 Review Later
#1
The Dartmouth Workshop of 1956, organized by John McCarthy, Marvin Minsky, and Claude Shannon, marked the formal establishment of AI.
#2
Machine learning divides into supervised learning (labeled data), unsupervised learning (unlabeled clustering), and reinforcement learning (reward signals).
#3
Artificial Neural Networks utilize multi-layer perceptrons trained through gradient descent and backpropagation to minimize loss functions.
#4
The Transformer architecture, published in 2017 ('Attention Is All You Need'), replaced recurrence with self-attention mechanisms in NLP.
#5
Computer vision utilizes Convolutional Neural Networks (CNNs) to process spatial pixel data for object recognition and autonomous systems.
#6
Alan Turing proposed the Turing Test in 1950 in his paper 'Computing Machinery and Intelligence' to evaluate whether a machine can exhibit indistinguishable human intelligence.
#7
Deep Blue, an IBM supercomputer utilizing heuristic tree search and specialized chess processors, defeated reigning World Chess Champion Garry Kasparov in a six-game match in May 1997.
#8
AlphaGo, developed by Google DeepMind, defeated 18-time world champion Go player Lee Sedol 4-1 in March 2016 using deep neural networks and Monte Carlo Tree Search.
#9
Generative Adversarial Networks (GANs), introduced by Ian Goodfellow in 2014, pit a generator network against a discriminator network in a zero-sum game framework.
#10
Overfitting occurs when a machine learning model learns noise and idiosyncratic details of training data, failing to generalize to unseen test datasets; it is mitigated by L1/L2 regularization and dropout layers.
#11
Recurrent Neural Networks (RNNs) process sequential temporal data, but suffer from vanishing gradient problems solved by Long Short-Term Memory (LSTM) units introduced by Hochreiter and Schmidhuber in 1997.
#12
Large Language Models (LLMs) rely on causal decoder-only transformer architectures with billions of parameters, pre-trained on token prediction and fine-tuned using Reinforcement Learning from Human Feedback (RLHF).
#13
Precision and Recall represent complementary classification evaluation metrics, combined mathematically into the harmonic mean F1-score to assess performance on imbalanced datasets.
#14
Gradient Descent updates network weights in the direction opposite to the gradient of the loss function, utilizing learning rates (eta) and optimizers such as Adam (Adaptive Moment Estimation).
#15
Natural Language Processing (NLP) tokenization converts raw textual sequences into subword vectors using algorithms such as Byte-Pair Encoding (BPE) or WordPiece.
Subject Specialist Commentary
Analytical perspective & practical exam advice from the Master10 academic board
Artificial intelligence enables computer software to perform tasks requiring human-like intelligence, such as recognizing images, making decisions, and translating text. Founded as an academic field at the 1956 Dartmouth Workshop, modern AI relies on machine learning to learn from data. Machine learning splits into supervised learning with labeled data, unsupervised learning for clustering unlabelled data, and reinforcement learning driven by rewards. Today's generative tools build upon artificial neural networks and transformer architectures that process language using self-attention mechanisms.
In UPSC Science and Technology and SSC exams, questions regularly test AI categories and practical applications. Avoid the common exam trap confusing algorithms: Convolutional Neural Networks specialize in image analysis, while Transformer models power modern Large Language Models. Remember that supervised learning depends on pre-labeled datasets, unlike unsupervised clustering. When studying for your test, note that overfitting describes a model memorizing training noise, which data scientists fix using regularization and dropout layers.
Related Knowledge Topics to Discover
Looking for more GK practice?
Explore 52,789+ questions across 65 General Knowledge categories.