Machine Learning: 2 Books in 1: An Introduction Math Guide for Beginners to Understand Data Science Through the Business Applications by Samuel Hack

Author:Samuel Hack [Hack, Samuel]
Language: eng
Format: azw3
Published: 2020-05-15T16:00:00+00:00

Types of Machine Learning

Supervised Machine Learning

The "supervised machine learning" is widely used in predictive big data analysis because they can assess and apply the lessons learned from previous iterations and interactions to new data set. These learning algorithms are capable of labeling all their current events based on the instructions provided to efficiently forecast and predict future events. For example, the machine can be programmed to label its data points as "R" (Run), "N" (Negative) or "P" (Positive). The machine-learning algorithm then labels the input data as programmed and gets the correct output data. The algorithm compares the production of its own with the "expected or correct" output, identifies potential modifications, and resolves errors to make the model more accurate and smarter. By employing methods like "regression", '' prediction", ''classification", and "boosting of ingredients" to properly train the learning algorithms, any new input data can be fed to the machine as "target" data set to assemble the learning program as desired. This jump-starts the analysis and propels the learning algorithms to create an "inferred feature", which can be used to generate forecasts and predictions based on output values for future events. Financial organizations and banks, for example, depend heavily on machine-learning algorithms to track credit card fraud and foresee the likelihood of a potential customer not making their loan payments on time.

Unsupervised Machine Learning

Companies often find themselves in a situation in which data sources are required to generate a labeled and categorized training data set are unavailable. In these conditions, the use of unsupervised machine learning is ideal. “Unsupervised learning algorithms” are commonly used to describe how the machine can produce "inferred features" to illustrate hidden patterns from an unlabeled and unclassified component in the stack of data. These algorithms can explore the data so that a structure can be defined within the data mass. Although the unsupervised machine learning algorithms are as effective as the supervised learning algorithms in the exploration of input data and drawing insights from it, the unsupervised algorithms are not capable of identifying the correct output. These algorithms can be used to define data outliers; to produce tailor-made product suggestions; to classify text topics using techniques such as "self-organizing maps”, "singular value decomposition" and "k-means clustering". Customer identification, for example, customers can be segmented into groups with shared shopping attributes and targeted with similar marketing strategies and campaigns. Consequently, unsupervised learning algorithms are very common in the online marketing industry.

Semi-Supervised Machine Learning

The "semi-supervised machine learning algorithms" are extremely flexible and able to learn from both “labeled” as well as “unlabeled” or raw data. These algorithms are a "hybrid" of supervised and unsupervised ML algorithms. Usually, the training data set consists of predominantly unlabeled data and a tiny portion of labeled data. The use of analytical methods such as the "forecast", "regression" and "classification" in combination with semi-controlled learning algorithms allows the computer to improve its accuracy in learning and training significantly. These algorithms are often used when the production of processed and labeled training data from the raw data set is highly resource-intensive and less cost-effective for the company.

Download

Copyright Disclaimer:
This site does not store any files on its server. We only index and link to content provided by other sites. Please contact the content providers to delete copyright contents if any and email us, we'll remove relevant links or contents immediately.