DEV Community

Cover image for Basic Terms in Machine Learning (Model Training)
John Johnson Okah
John Johnson Okah

Posted on

Basic Terms in Machine Learning (Model Training)

As the AI hype grows, machine learning is making its way into every recent innovative product. Therefore, understanding the basic concepts and terminologies of machine learning becomes increasingly essential for anyone looking to get started with or deepen their knowledge of machine learning.
Machine learning is a powerful technique that enables computers to learn from data and make predictions or decisions without being explicitly programmed.

Machine learning concepts

Below are brief explanations of some fundamental terminologies in machine learning.

Basic Terms in Machine Learning

  • Artificial Intelligence: Artificial Intelligence (AI) is a technology that enables computers and machines to simulate human intelligence and problem-solving capabilities.
  • Machine Learning: Machine Learning (ML) is a subset of artificial intelligence that involves training algorithms to learn from data and make predictions or decisions without being explicitly programmed to perform a specific task.
  • Neural Network: A neural network is a type of machine learning model that is inspired by the structure and function of the human brain.
  • Deep Learning: Deep learning is a subset of machine learning that uses artificial neural networks with multiple layers to model complex relationships between inputs and outputs.
  • Model: A model is a simplified representation of a real-world process or system that is created to make predictions or decisions based on input data.
  • Dataset: A dataset is a structured collection of data that is organized and stored for easy access and analysis. They are sets of examples or observations that are used to train, validate or test machine learning models.
  • Prediction: The prediction is the result/output from the model. It is the estimation based on input data.
  • Parameters: Parameters are internal variables of a machine learning model that are learned from the training data to capture the relationships and patterns within the data.
  • Training: This is the process of teaching a model to make predictions or decisions by exposing it to input data and adjusting its internal parameters based on the observed outcomes.
  • GPU: Graphics Processing Unit (GPU) is a specialized hardware components that excel at parallel processing and has become indispensable in accelerating machine learning tasks.


At the core of machine learning is the process of training models - algorithms that can recognize patterns and extract insights from data.
Here is an oversimplified process of machine learning model training:

first, you collect and clean relevant data, then pick an algorithm suited to the task. The algorithm analyzes the data to learn patterns, and the model is tested and refined to improve its accuracy. Finally, the trained model can be used to make predictions on new data.

machine learning model training process

Now, let's look at some terms in training a machine learning model.

Basic Terms in Model Training

  • Pre-trained models: These are models that have been trained on a large dataset to perform a specific task, such as image classification, natural language processing, etc., before being made available for use by others.
  • Transfer learning: This is a machine learning technique that leverages pre-trained model knowledge to be applied to a different but related task (by fine-tuning).
  • Fine-tuning: It is a process in transfer learning where a pre-trained model is further trained on a new, smaller dataset for a specific task.
  • Overfitting: This occurs when a model learns the training data too well, capturing noise and random fluctuations rather than the underlying patterns. It leads to poor generalization performance on new, unseen data.
  • Training set, validation set, and test set: These are different portions of the dataset for evaluation. The training set is used to train the model, the validation set is used to fine-tune hyperparameters and evaluate performance during training, and the test set is used to assess the model's final performance on unseen data.
  • Hyperparameters: These are other parameters that define how the model is trained. For example, defining how long we train for, or what learning rate. They are external configuration settings set by the engineer or researcher.
  • Epoch: An epoch refers to one complete pass through the entire training dataset during the training of a model. Multiple epochs are typically required. On each epoch, the parameters are updated based on the gradient of the loss function.
  • Metric: A metric is a function that measures the quality of the model's predictions using the validation set. A metric is defined for human consumption, easy for you to understand.
  • Loss: This is also known as the loss function or cost function. It is a mathematical function that quantifies the difference between the model's predictions and the actual outcomes (labels) in the training data. Loss is used to update weights automatically, easy for stochastic gradient descent to use.
  • Architecture
  • Stochastic Gradient Descent: SGD is an optimization algorithm used for training models by updating the parameters iteratively with small batches of data.
  • Architecture: This is the overall structure design of a model. The architecture of a neural network defines its overall structure, including the arrangement of layers, the number of neurons, and the connections between neurons.

Having gone through this ML jargon, you should have a solid grasp of the core terminology used in machine learning model training. This knowledge will prepare you to dive deeper into more advanced ML concepts.

Top comments (0)