Learning Using Python: How to Build Applied Machine Learning Solutions
Author: Reed Cartwright
ASIN: B09W79N1N8
Publisher finelybook 出版社: Independently published (March 22, 2022)
Language 语言: English
Print Length 页数: 457 pages
ISBN-13: 9798437574973
Book Description
By finelybook
Many market professionals consider unsupervised learning the following frontier in artificial intelligence, one that might hold the secret to the holy grail in AI study, the so called general expert system. Considering that the majority of the world’s data is unlabeled, traditional supervised knowing can not be applied; this is where without supervision knowing can be found in. Without supervision understanding can be related to unlabeled datasets to find significant patterns hidden deep in the data, patterns that may be near difficult for humans to uncover.
Author gives sensible understanding on exactly how to use not being watched understanding making use of two basic, production prepared Python structures scikit learn and also TensorFlow utilizing Keras. With the hands on instances and code supplied, you will determine hard to find patterns in information as well as get much deeper business understanding, find anomalies, execute automated function design as well as choice, and generate artificial datasets. All you require is programming as well as some equipment learning experience to get started.
Contrast the strengths and also weaknesses of the various machine finding out strategies: monitored, without supervision, and also support knowing
Establish and handle an equipment learning task end to end every little thing from information procurement to building a model and also executing a solution in production
Use dimensionality reduction algorithms to uncover one of the most pertinent details in data and also build an abnormality detection system to capture bank card fraud
Apply clustering algorithms to sector users such as lending borrowers right into distinctive as well as homogeneous groups
Usage autoencoders to perform automated attribute engineering and also option
Integrate supervised and also not being watched knowing formulas to develop semi monitored remedies
Build motion picture recommender systems using restricted Boltzmann devices
Generate synthetic photos utilizing deep idea networks as well as generative adversarial networks
Perform clustering on time collection data such as electrocardiograms
Check out the successes of without supervision learning to day and its appealing future
