Additional Terminology Flashcards
What is Backpropagation Through Time (BPTT)?
A method to train recurrent neural networks by unfolding them in time.
What is the Bag-of-Words model in NLP?
A representation of text that disregards grammar and word order but keeps word frequency.
What are biases in machine learning?
Systematic errors introduced in models due to assumptions in data or algorithms.
What is selection bias in datasets?
A bias introduced when the sample is not representative of the population.
What is confirmation bias in data interpretation?
The tendency to favor information that confirms one’s preconceptions.
What is historical bias in datasets?
Bias introduced due to outdated or non-representative historical data.
What is labelling bias in datasets?
Incorrect or inconsistent data labeling that affects model performance.
What is linguistic bias in datasets?
A bias that occurs when certain linguistic groups are favored over others in data.
What is sampling bias in datasets?
Bias caused by a non-random selection of data samples.
What is the role of a dataset in machine learning?
A collection of data used to train and evaluate machine learning models.
What is deep learning?
A subset of machine learning using neural networks with many layers.
What is a graphical processing unit (GPU) in deep learning?
A hardware unit that accelerates the computation of deep learning models.
What is hyperparameter tuning in machine learning?
The process of adjusting model settings to optimize performance.
What is a large language model (LLM)?
A neural network model trained on massive amounts of text data for language tasks.
What is latency in computing?
The delay before a transfer of data begins following an instruction.
What is long short-term memory (LSTM)?
A type of recurrent neural network that can learn long-term dependencies.
What is a loss function in machine learning?
A function that measures the difference between the predicted and actual outputs.
What is the memory cell state in LSTM?
The internal state of the LSTM that stores information across time steps.
What is natural language processing (NLP)?
A field of AI focused on the interaction between computers and human language.
What is discourse integration in NLP?
The process of linking sentences together in a coherent text.
What is lexical analysis in NLP?
The process of breaking text into tokens, such as words or phrases.
What is pragmatic analysis in NLP?
Understanding the intended meaning of words in context, beyond literal meaning.
What is semantic analysis in NLP?
The process of understanding the meaning of words and sentences.
What is syntactical analysis (parsing) in NLP?
The process of analyzing sentence structure to understand grammar.
What is natural language understanding (NLU)?
A subset of NLP that focuses on machine comprehension of text meaning.
What is pre-processing in machine learning?
Data preparation steps like cleaning and transforming before model training.
What is a recurrent neural network (RNN)?
A neural network model that processes sequential data by maintaining a memory of past inputs.
What is the self-attention mechanism?
A method used in transformers to focus on different parts of the input sequence for better context understanding.
What is synthetic data?
Data artificially generated rather than obtained by direct measurement.
What is a tensor processing unit (TPU)?
A hardware accelerator designed by Google for efficient deep learning computation.
What is a transformer neural network (transformer NN)?
A model architecture designed to handle sequential data using self-attention mechanisms.
What is the vanishing gradient problem?
A problem where gradients diminish during backpropagation, making it difficult to train deep neural networks.
What are weights in neural networks?
The parameters in the network that are adjusted during training to minimize the loss.
