Table of Contents

Models Supported

Hardik Chheda Updated by Hardik Chheda

Tellius supports following four type of machine learning models with lots of model in each of these sections. 

  • Regression
  • Classification
  • Time series regression
  • Clustering

Regression

ML models for regression predicts a numeric value. In this model, Tellius uses historical data to build models and predict the attributes, for example, traffic, home price, inventory, and so on.

Supported algorithms

For training regression models, Tellius ML models use the following industry-standard learning algorithm: 

  • Linear Regression: This is an approach for modeling a relationship between the dimensions or features and one or more measures.
  • Tree Regression:  This is a decision support tool that uses a tree-like graph or model of decisions and their possible consequences.
  • Linear Regression with Regularization: This is an approach same as Linear Regression with addition process of introducing additional information to prevent overfitting, which is one of the most common tasks to fit a "model" to a set of training data.
  • GLM Regression: The generalized linear model (GLM) is a flexible generalization of conventional linear regression that allows the linear model to be related to the response variable through link function and the magnitude of the variance of each measurement to be the function of its predicted value. You can perfrom the GLM regression on binary outcome data, count data, probability data, proportion data and many other data types.
Examples of Regression model:
  • "What will the temperature be in Seattle tomorrow?"
  • "For this product, how many units will sell?"
  • "What price will this house sell for?"

Classification

In this model, Tellius uses historical data/ patterns to perform classification. Tellius supports binary classification, multiclass variable classification models.

Supported algorithms

For training classification models, Tellius ML models use the following industry-standard learning algorithm: 

  • Decision Tree Classifier: Decision Tree Classifier is a simple and widely used classification technique, which applies an idea to solve the classification problem.
  • Naive Bayes: Naive Bayes is a family of simple probablistic classifier based on applying Bayes theorm with strong (naive) independence assumptions between the features.
  • Logistic regression: Logistic regression is simple technique where the dependent variable is categorical.
  • Artificial Neural Network Classifier: An artificial neural network is an interconnected group of nodes, akin to the vast network of neurons in a brain. Here, each node represents an artificial neuron and an arrow represents a connection from the output of one neuron to the input of another.
  • Random Forest Classifier: Random Forest Classifier are an ensemble learning method for classification, regression, and other tasks, that operate by constructing a multitude of decision trees at training time and outputting the class that is the mode of the classes (classification) or mean prediction (regression) of the individual trees. 
  • GBT Classifier: Gradient-Boosted Trees (GBTs) is a learning algorithm for classification. It supports binary labels, as well as both continuous and categorical features.
  • GLM Classifier
  • Linear Support Vector Machine 
Examples of Classification model
  • "Is this email spam or not spam?"
  • "Will the customer buy this product?"
  • "Is this review written by a customer or a robot?"  

Time series regression

In this model, Tellius performs the forecasting for time series data, for example, predicting stock prices, retail sales for next 30 to 60 days based on past trends.This model is similar to the regression model, except that you can perform the time base predictive analytics of your data. 

Supported algorithms

For training regression models, Tellius ML models use the following industry-standard learning algorithm: 

  • ETS Regression: ETS (Error, Trend, Seasonal/Exponential Smoothing) provides an automatic way of selecting best Exponential Smoothing method from 30 separate models in ETS framework.
  • ARIMA Regression: ARIMA Regression model can be considered as a special type of regression model, in which the dependent variable (dimension/ feature) has been stationed and the independent variables (measures) are all lags of the dependent variable and/or lags of the errors.
  • STLM Regression: Applies a STL decomposition  (seasonal, trend, and error components using Loess - Loess is a non-linear regression technique)
  • NAIVE: The NAIVE algorithm is a classification technique based on Bayes' Theorem with an assumption of independence among features. In simple terms, a Naive Bayes classifier assumes that the presence of a particular feature in a class is unrelated to the presence of any other feature.
  • Seasonal NAIVE
  • DRIFT: DRIFT means that the statistical properties of the target variable, which the model is trying to predict, change over time in unforeseen ways. This causes problems because the predictions become less accurate as time passes.
  • TBATS: A TBATS algorithm could be considered as a time series decomposition method that allows multiple complex seasonalities to be incorporated simultaneously. For example, there may be a weekly seasonal component and a monthly seasonal component which both need to be incorporated into the forecast model.
  • Neural Networks: Neural Network (NN) is an information-processing paradigm that is inspired by the way biological nervous systems, such as the brain, process information. 
Examples of Time Series Regression model
  • Rate of unemployment for last 10 years
  • Rate of price inflation measured by quarterly percentage change in the price index at an annual rate 

Clustering

Cluster analysis or clustering is the task of grouping a set of objects in such a way that objects in the same group (called as a cluster) are more similar (in some way or another) to each other than to those in other groups (clusters).  

In this model, Tellius performs the categorization of multiple clusters and use that clusters to detect unusual patterns or similar segments of data.

Supported algorithms

For training regression models, Tellius ML models use the following industry-standard learning algorithm: 

  • K Clustering: K Clustering is a method of vector quantization, originally from signal processing, that is popular for cluster analysis in data mining.
  • Bisecting K-Means: Bisecting k-Means is like a combination of k-Means and hierarchical clustering. It starts with all objects in a single cluster. The pseudocode of the algorithm is displayed below: Basic Bisecting K-means Algorithm for finding Kclusters 
Examples of Clustering model
  • Discovering distinct groups in customer bases, and then using that knowledge to develop targeted marketing programs
  • Identifying groups of motor insurance policy holders with a high average claim cost
  • Identifying groups of houses according to their house type, value, and geographical location 

Recommender Systems

A recommender system or a recommendation system (sometimes replacing "system" with a synonym such as platform or engine) is a subclass of information filtering system that seeks to predict the "rating" or "preference" that a user would give to an item. 

You specify Rating column, User column and item column in the feature transform menu.

How did we do?

What is Point-n-Click ML

Contact