Types of Machine Learning Algorithms

Types of Machine Learning Algorithms.

On this article you will get to know the Types of Machine Learning Algorithms.

  1. Linear Regression.
    This algorithm uses variables to estimate the cost of real values.It exists in two types.
    Simple Linear Regression which consists of independent variables.
    Multiple Linear Regression which consists of more than one variable.
    Using an example for more vivid understanding,lets say
  2. Logistic Regression.
    It uses a classification logarithm that is used to predict the probability of the pattern of binary values(0/1) by fitting them into a logic function.The algorithm reviews the probability of maximising the number of the binary values provided as compared to the minimum error that are likely to occur.
  3. Decision Tree.
    This algorithm is mainly used in supervised learning which helps in classifying items from the highest in the hierarchy to the lowest.It works by differentiating data and placing similar data in the same group.
  4. Support Vector Machine(SVM).
    This is an algorithm that is used to plot data as the number of features contained in the data vs the value of each feature(coordinates).The data can be separated using a black line(hyperplane).It is mainly used in classifying data into classes,balancing data,cross validation and automatic model selection.
  5. Naive Bayes.
    It is used for large amount of datasets and uses the Bayes theorem to show how the presence of specific features are unrelated.
  6. K-Nearest Neighbours.
    This is an algorithm that is used to solve problems by classification method and regression.It is mostly used for classification of large amounts of data.The majority cases are grouped into K neighbours and the most common are measured by distance function.Some of the distance functions are  Euclidean,Manhattan,Minkowski and Hamming.They are used for both continous functions and categorical variables.
    KNN is a very expensive algorithm.If the variables are not optimized they become bias and the process focuses on the pre-processing stage more before engaging in KNN.
  7. K-Means
    The centroids are gathered into a cluster and numbered as letter K until an optimal centroid is attained.
  8. Random Forest
    It is used for tasks such as classification and regression.It is simply a group of decision trees that works by splitting nodes with random features.The ensemble learning technique groups complex problems and provides solutions to them.
  9. Dimensionality Reduction Algorithms.
    This is the convertion of data,both numerical and letters,into a more simplified and understandable form.The most commonly used methods for data dimensionality include Ration of missing values,Low variance in the column values,High correlation between two columns,principal component analysis,backward feature elimination etc.
  10. Gradient Boost and Adaboost.

Gradient boosting refers to the process of finding a solution to additive modelling problems.It works by combining the best futuristic model with the best past model to attain a model with minimal error.Adaptive boosting(Adaboost) is an algorithm that uses the ensemble method to assign higher weights to the classification labelled incorrectly.

Areas where Machine learning is used.

  1. Self-driving cars.
  2. Speech translation.
  • Face recognition.
  1. Social media analysis
  2. Fraud detection
  3. Medical diagnosis and predictions
  • Product recommendation.



Top best Website design Company in Kenya

Best Point of sale system in Kenya

Mobile app development in kenya

Top best software development in Kenya

Best school management System in Nairobi,kenya