ML Algorithms vs ML Models
- [The University of Chicago - Vivian Wu]
- Overview
In the context of machine learning (ML), a ML algorithm refers to the mathematical procedure or set of rules used to analyze data and identify patterns, ML model is the concrete output of that algorithm, essentially a program that can make predictions based on the learned patterns from the data; in simpler terms, the algorithm is the recipe, and the model is the finished dish created using that recipe.
Key Differences:
- Function: An ML algorithm defines the process for learning from data, while an model is the actual representation of that learned knowledge, ready to be used for predictions.
- Output: An algorithm produces a model as its output after being applied to data.
- Flexibility: Algorithms can be applied to different datasets, while a specific model is tailored to the data it was trained on.
Example:
- Algorithm: Linear regression - a mathematical formula that calculates the relationship between variables.
- Model: A specific linear regression equation with calculated coefficients, generated after training on a particular dataset.
- ML Models
The model serves as the underlying core component of ML, representing the link between inputs and outputs to produce accurate and fresh data. It is trained on data sets to identify underlying patterns and produce accurate results.
After training, the model is tested to determine whether it can provide fresh and accurate data; if the test is successful, it is used in real-world applications.
Let us take an example to understand this further. You want to build a model that takes into account characteristics such as age, body mass index (BMI), and blood sugar levels to identify whether a person has diabetes.
We had to first compile a dataset of diabetes patients and related health indicators. The algorithm uses a dataset of diabetic patients and considers their health indicators to analyze patterns and relationships in the data and produce accurate results. It identifies potential relationships between outcomes (diabetes status) and input characteristics (blood glucose levels, BMI, and age).
After training, the model can use information such as blood sugar levels, weight and age to predict whether a new patient has diabetes.
- ML Algorithms
Train models using algorithms that learn hidden patterns from data, predict outputs, and improve performance with experience. It is an important component of ML because it powers the learning process and affects the accuracy and effectiveness of the model.
The training data set consists of input data and associated output values. Once patterns and associations in the data are identified, a variety of mathematical and statistical techniques are used to determine the underlying relationships between inputs and outputs.
For example, when we have a dataset of animal photos and their matching species labels, we need to train a ML model to identify the species of the animals in the photos. Convolutional Neural Networks (CNN) can be used for this purpose.
The CNN method breaks incoming visual data into multiple layers of mathematical operations to identify features such as edges, shapes, and patterns. These features are then used to classify the image into one of the species categories.
However, there are several alternatives, including decision trees, logistic regression, k-nearest neighbors, etc. The data set provided and the problem that must be solved determine your algorithm.
- [Machine Learning Cheat Sheet]
- Some ML Techniques and Algorithms
Some key machine learning techniques and algorithms include: linear regression, logistic regression, support vector machines (SVM), Naive Bayes, decision trees, random forests, K-Nearest Neighbors (KNN), clustering, dimensionality reduction, gradient boosting, and AdaBoost.
The choice of which algorithm to use depends on the specific data and problem you're trying to solve, factors like data size, quality, and desired outcome playing a significant role.
Here are some ML algorithms:
- Logistic regression: A ML technique that is good for binary classification problems. It uses a logistic function at its core.
- Decision tree: A ML technique that uses rules and conditions to solve classification problems. It involves dividing input data into two or more homogeneous data sets based on defining attributes.
- Support vector machine: A ML method that uses statistical theory to solve fitting accuracy and generalization problems. It is used in pattern recognition, information security, and data fitting.
- Naive Bayes: A ML algorithm that is used when the output variable is discrete. It is driven by the Bayes Theorem.
- Random forest: A ML process that consists of many decision trees. A decision tree is a tree-like structure where each internal node represents a test on the input attribute.
- Clustering: A ML technique that involves grouping data points. It is an unsupervised learning method and a famous technique for statistical data analysis.
- Hyperparameters: An integral part of machine learning code that lets you control the code without directly modifying it.
- Gradient descent: A famous optimization technique that is used in machine learning and in deep learning. Its main purpose is to minimize the cost function.
- Some ML Models
Machine learning models are computer programs that use algorithms to recognize patterns in data or make predictions. They are trained using labeled, unlabeled, or mixed data.
Here are some examples of machine learning models:
- Support Vector Machines (SVM): A popular model used for validation because it can maximize the margin between data points of different classes
- Logistic regression: A statistical model that predicts the class of the dependent variable from a set of independent variables. It is a popular method for solving binary classification problems
- Decision trees: A supervised learning technique used for classification and regression. It is a tree-structured classifier where internal nodes represent dataset features, branches represent decision rules, and every leaf node represents the outcome
- Naive Bayes: A set of supervised learning algorithms used to create predictive models for binary or multi-classification tasks
- Reinforcement learning: A behavioral modeling technique where the model learns through a trial and error mechanism
- Neural networks: A subset of machine learning that is made up of artificial neurons and designed to resemble the human brain structure and working
- K Means Clustering: An unsupervised learning algorithm used to categorize unlabeled data
- Deep learning: A subset of machine learning that uses multilayered neural networks to simulate complex decision-making
[More to come ...]
