Machine Learning for Data Science Handbook: Data Mining and Knowledge Discovery Handbook by Lior Rokach & Oded Maimon & Erez Shmueli

Author:Lior Rokach & Oded Maimon & Erez Shmueli
Language: eng
Format: epub
ISBN: 9783031246289
Publisher: Springer International Publishing

4 Approximate Computing (AC) and Cloud Big Data Analytics

Approximate computing or In-exact computing is a technique for trading off result accuracy with speed and energy. It has been successfully used in many domains ranging from machine learning to financial data analysis. Big data is a very strong and important target for AC as big data does not always require exact analytics output but needs a summary of the output. Losing a few data items of big data while performing analytics will not impact the final result of mining and analytics. A very small change in the data often does not have the caliber to change the meaning of the predicted value or shift its importance [32]. Some recent articles [32, 7] state the techniques and applications of AC used in big data and machine learning perspective in the cloud and related distributed platforms. The remainder of this section discusses the usage and applications of AC in big data mining and machine learning in cloud environments. Figure 3 gives an idea of the architecture for using AC techniques atop cloud data mining or analytics frameworks. The major techniques are task/job skipping, memoization, and memory skipping. Data sampling can also be applied to cut down unnecessary data and use only a small subset of it for further processing without much change in the mining output.

Fig. 3 The general architecture for using AC in Cloud Big Data Analytics

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.