Reverse Hypothesis Machine Learning

A Practitioner's Perspective

verfasst von: Parag Kulkarni

Verlag: Springer International Publishing

Buchreihe : Intelligent Systems Reference Library

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Einloggen, um Zugang zu erhalten

Über dieses Buch

This book introduces a paradigm of reverse hypothesis machines (RHM), focusing on knowledge innovation and machine learning. Knowledge- acquisition -based learning is constrained by large volumes of data and is time consuming. Hence Knowledge innovation based learning is the need of time. Since under-learning results in cognitive inabilities and over-learning compromises freedom, there is need for optimal machine learning. All existing learning techniques rely on mapping input and output and establishing mathematical relationships between them. Though methods change the paradigm remains the same—the forward hypothesis machine paradigm, which tries to minimize uncertainty. The RHM, on the other hand, makes use of uncertainty for creative learning. The approach uses limited data to help identify new and surprising solutions. It focuses on improving learnability, unlike traditional approaches, which focus on accuracy. The book is useful as a reference book for machine learning researchers and professionals as well as machine intelligence enthusiasts. It can also used by practitioners to develop new machine learning applications to solve problems that require creativity.

Inhaltsverzeichnis

Frontmatter

Building Foundation: Decoding Knowledge Acquisition

Frontmatter

Chapter 1. Introduction: Patterns Apart

Abstract

Pattern is such a beautiful word and is an interesting data representation. There is pattern everywhere and we have firm belief that we learn through patterns. Pattern has taught us almost everything. There is pattern in natural events, there is pattern in behavioral aspect and there is pattern in even obvious and nonobvious behaviors.

Parag Kulkarni

Chapter 2. Understanding Machine Learning Opportunities

Abstract

Machine learning has developed over the years. Right from simple rote learning, we have used different paradigms of machine learning to handle some very complex learning tasks. It has converted many so-called fictions into reality. Traditionally, machine learning was confined to acquiring more information and retrieving the information timely. It was more like information acquisition to information retrieval. Let us take an example of a student who claims that he learned deep learning. What does that mean? Does he understand the concept or just few terminologies? Can he apply these concepts? Can he expand these concepts?

Parag Kulkarni

Chapter 3. Systemic Machine Learning

Abstract

As we have discussed in last two chapters, learning in isolation has its own limitations. One of the core strengths of human is to work in context and associate with different scenarios to build context. In different contexts, same information can lead to different out come and can help in producing different possibilities. In linguistic sense same word may have different meanings as per context.

Parag Kulkarni

Chapter 4. Reinforcement and Deep Reinforcement Machine Learning

Abstract

Data-driven learning is a very strong concept. This concept is chased and converted into wonderful applications. Whole stream of Data Engineering and Data Sciences emerged out of that. The data is collected from various sources. It is collected from big hospitals, data repositories, from cookies running in your machines, intelligent applications running in your devices. It can be crowd sourcing, intelligent crowd sourcing, cognitive data sharing, and mind sharing. The intelligent data collection and usage delivered applications which proved that crowd is always ahead of everyone. It has gone from power of two, to power of few, to power of crew, to power of many.

Parag Kulkarni

Learnability Route: Reverse Hypothesis Machines

Frontmatter

Chapter 5. Creative Machine Learning

Abstract

While paradigm of knowledge-based intelligence has focus on using and acquiring knowledge, exploratory learning has focus on more learning and hence driven by new scenarios along with existing knowledge. Recently there is all focus on exploratory learning to cater with complexity of real life problems. Knowledge acquisition-based learning has its own limitations. These limitations range from inability in handling slightly different scenario to failing to create some real nonobvious solution. Exploration is definitely an interesting concept, but knowledge acquisition and exploratory learning work perfectly in some real life scenarios and bring wonderful results.

Parag Kulkarni

Chapter 6. Co-operative and Collective Learning for Creative ML

Abstract

Crowdsourcing is a form of collective learning. Learning collectively refers to getting together to solve problem and learn from each other’s experience. One of the basic forms is collecting information or solutions from more than one intelligent agent. This is referred to as crowdsourcing. Ensemble learning is a concept where there is an ensemble of learners and based on algorithm it is used. Ensemble learner has more than one learner—even boosting is a concept where more than one learner is learning. Sometimes a set of weak classifiers is used for building strong classifier.

Parag Kulkarni

Chapter 7. Building Creative Machines with Optimal ML and Creative Machine Learning Applications

Abstract

Why do we need creative machine learning? Do you need to make machines more intelligent? We have ample applications of Forward hypothesis Machine. Right from driverless cars to all different machines used in day to day life are using FHM. Then why we need reverse hypothesis machine? Deep learning and allied technologies have enabled solving of complex problems. Then why is the need of creative ML? We will try to justify the need of creative ML and explain a few applications of it. Creative learning is required in many cases those range from art, science, and innovation to contribution to growth. Creativity is associated with learning in many ways. One cannot be absolute creative without learning. Today what is surprising may not be surprising tomorrow.

Parag Kulkarni

Chapter 8. Conclusion—Learning Continues

Abstract

Learning is necessary for every intelligent creature and in modern era every intelligent system. Can we term knowledge acquisition same as learning?—is the question for discussion here. Learning has many facets and there can be difference in opinion—about what to be called learning and what should not be called learning.

Parag Kulkarni

Backmatter

Titel: Reverse Hypothesis Machine Learning
verfasst von: Parag Kulkarni
Verlag: Springer International Publishing
Electronic ISBN: 978-3-319-55312-2
Print ISBN: 978-3-319-55311-5
DOI: https://doi.org/10.1007/978-3-319-55312-2