MindLink

Matching Pairs

Played 1

Create copy

About this activity

Connect concepts, master Machine Learning.

MindLink: Create Matching Pairs is an interactive challenge for 1st year AIML students. Participants will match key terms with their correct definitions, covering Machine Learning, Deep Learning, and Reinforcement Learning concepts across multiple rounds.

A fun and engaging way to strengthen your ML knowledge, sharpen recall, and test conceptual understanding.

Created by

LEELANANDHA KISHORE K S BTech_AIML

India

Download the paper version to play

Top 10 results

There are still no results for this game. Be the first to stay in the ranking! Log in to identify yourself.

Make your own free game from our game creator

Compete against your friends to see who gets the best score in this game

Make challenge

		%
	Anonymous	%
		%
		%

Print your game

MindLink


Matching Pairs MindLinkOnline version Connect concepts, master Machine Learning. MindLink: Create Matching Pairs is an interactive challenge for 1st year AIML students. Participants will match key terms with their correct definitions, covering Machine Learning, Deep Learning, and Reinforcement Learning concepts across multiple rounds. A fun and engaging way to strengthen your ML knowledge, sharpen recall, and test conceptual understanding. by LEELANANDHA KISHORE K S BTech_AIML 1 Training Data 2 Clustering 3 Feature 4 Regression 5 Classification 6 Unsupervised Learning 7 Label 8 Supervised Learning Learning with unlabeled data Predicting continuous values Grouping similar data points Output variable in supervised learning Learning with labeled data Data used to train the model Predicting categories Input variable in a model 1 Underfitting 2 F1-Score 3 Precision 4 Accuracy 5 Overfitting 6 Cross-Validation 7 Recall 8 Confusion Matrix Table showing true vs predicted labels Model fails to capture patterns Harmonic mean of precision and recall Technique to assess model performance True positives / Predicted positives True positives / Actual positives Correct predictions / Total predictions Model fits training data too closely 1 Stop Words 2 Dimensionality Reduction 3 Tokenization 4 Feature Selection 5 Standardization 6 One-Hot Encoding 7 PCA 8 Normalization Scaling to zero mean & unit variance Scaling features to a specific range Breaking text into words or tokens Choosing important variables Principal Component Analysis Reduce number of features Convert categorical to binary vectors Common words removed during text processing 1 Dropout 2 Activation Function 3 Batch 4 Perceptron 5 Epoch 6 Gradient Descent 7 Perceptron 8 Neural Network Regularization technique to prevent overfitting One complete pass over training data Optimization algorithm Simplest type of neural network Adds non-linearity to network Model inspired by human brain Subset of data processed at a time Algorithm to update weights 1 Time Series 2 GRU 3 Convolution 4 Sequence Data 5 Convolution 6 LSTM 7 RNN 8 CNN Reducing spatial dimensions Long Short-Term Memory network Data with order dependencies Gated Recurrent Unit Feature extraction operation Sequential data over time Recurrent Neural Network Convolutional Neural Network 1 BERT 2 Embedding 3 Self-Attention 4 Transformer 5 Encoder 6 Decoder 7 NLP 8 Attention Converts input to representation Natural Language Processing Uses attention mechanism Focus on important parts of input Attention within same sequence Pretrained transformer model Representation of data in vector space Generates output from representation 1 Agent 2 Exploration 3 Action 4 Reward 5 State 6 Environment 7 Policy 8 Episode Strategy to select actions Learner or decision-maker World in which agent interacts Sequence of states, actions, rewards Feedback signal Current situation of agent Move taken by agent Trying new actions 1 Bellman Equation 2 SARSA 3 Q-Learning 4 Policy Gradient 5 Exploitation 6 Markov Decision Process 7 Discount Factor 8 Value Function Using known best actions Framework for RL problems Value-based RL algorithm Computes value functions Estimates expected reward On-policy RL algorithm Future reward weighting Optimizes policy directly 1 Cross-Entropy 2 Learning Rate 3 Convergence 4 Batch Normalization 5 Weight Initialization 6 Softmax 7 Hyperparameter 8 Regularization When training stabilizes Parameter set before training Loss function for classification Reduces overfitting Step size in gradient descent Normalizes layer inputs Converts logits to probabilities Starting weights in neural nets 1 Overfitting 2 Deep Learning 3 Fine-Tuning 4 Reinforcement Learning 5 Transfer Learning 6 Attention Mechanism 7 NLP 8 Feature Engineering Using pretrained models Adjusting pretrained model on new data Creating new features Processing human language Learning via rewards Focus on important input parts Neural network-based learning Model memorizes training data

MindLink

Matching Pairs

About this activity

Created by

Download the paper version to play

Top Games

Matching Pairs

The Bill of Rights

Matching Pairs

Cranial Nerves

Matching Pairs

Cell Cycle Memory

Matching Pairs

Balance Sheet & P & L

Matching Pairs

appendicular skeleton

Matching Pairs

MindLinkOnline version

by LEELANANDHA KISHORE K S BTech_AIML

Training Data

Clustering

Feature

Regression

Classification

Unsupervised Learning

Label

Supervised Learning

Underfitting

F1-Score

Precision

Accuracy

Overfitting

Cross-Validation

Recall

Confusion Matrix

Stop Words

Dimensionality Reduction

Tokenization

Feature Selection

Standardization

One-Hot Encoding

PCA

Normalization

Dropout

Activation Function

Batch

Perceptron

Epoch

Gradient Descent

Perceptron

Neural Network

Time Series

GRU

Convolution

Sequence Data

Convolution

LSTM

RNN

CNN

BERT

Embedding

Self-Attention

Transformer

Encoder

Decoder

NLP

Attention

Agent

Exploration

Action

Reward

State

Environment

Policy

Episode

Bellman Equation

SARSA

Q-Learning

Policy Gradient

Exploitation