scholarly journals EXPLORING THE STACKING STATE-SPACE

2002 ◽  
Vol 11 (02) ◽  
pp. 267-282 ◽  
Author(s):  
AGAPITO LEDEZMA ◽  
RICARDO ALER ◽  
DANIEL BORRAJO
Keyword(s):  
Machine Learning ◽  
State Space ◽  
Linear Response ◽  
Search Space ◽  
Learning Systems ◽  
Hill Climbing ◽  
Machine Learning Techniques ◽  
Learning Techniques ◽  
Inductive Machine Learning ◽  
Simple Search

Nowadays, there is no doubt that machine learning techniques can be successfully applied to data mining tasks. Currently, the combination of several classifiers is one of the most active fields within inductive machine learning. Examples of such techniques are boosting, bagging and stacking. From these three techniques, stacking is perhaps the less used one. One of the main reasons for this relates to the difficulty to define and parameterize its components: selecting which combination of base classifiers to use, and which classifier to use as the meta-classifier. One could use for that purpose simple search methods (e.g. hill climbing), or more complex ones (e.g. genetic algorithms). But before search is attempted, it is important to know the properties of the search space itself. In this paper we study exhaustively the space of Stacking systems that can be built by using four base learning systems: C4.5, IB1, Naive Bayes, and PART. We have also used the Multiple Linear Response (MLR) as meta-classifier. The properties of this state-space obtained in this paper will be useful for designing new Stacking-based algorithms and tools.

scholarly journals WHAT PERSONALISATION CAN DO FOR YOU! OR, HOW TO DO RACIAL DISCRIMINATION WITHOUT RACE

2021 ◽  
Author(s):  
Scott Wark ◽  
Thao Phan
Keyword(s):  
United States ◽  
Machine Learning ◽  
Asian American ◽  
Language Use ◽  
The United States ◽  
Machine Learning Techniques ◽  
Life And Death ◽  
Demographic Group ◽  
Learning Techniques ◽  
Inductive Machine Learning

Between 2016 and 2020, Facebook allowed advertisers in the United States to target their advertisements using three broad “ethnic affinity” categories: “African American,” “U.S.-Hispanic,” and “Asian American.” This paper uses the life and death of these “ethnic affinity” categories to argue that they exemplify a novel mode of racialisation made possible by machine learning techniques. These categories worked by analysing users’ preferences and behaviour: they were supposed to capture an “affinity” for a broad demographic group, rather than registering membership of that group. That is, they were supposed to allow advertisers to “personalise” content for users depending on behaviourally determined affinities. We argue that, in effect, Facebook’s ethnic affinity categories were supposed to operationalise a “post-racial” mode of categorising users. But the paradox of personalisation is that in order to apprehend users as individuals, platforms must first assemble them into groups based on their likenesses with other individuals. This article uses an analysis of these categories to argue that even in the absence of data on a user’s race—even after the demise of the categories themselves—users can still be subject to techniques of inclusion or exclusion for discriminatory ends. The inductive machine learning techniques that platforms like Facebook employ to classify users generate “proxies,” like racialised preferences or language use, as racialising substitutes. This article concludes by arguing that Facebook’s ethnic affinity categories in fact typify novel modes of racialisation today.

scholarly journals Genetic programming for region detection, feature extraction, feature construction and classification in image data

2020 ◽  
Author(s):  
Andrew Lensen ◽  
Harith Al-Sahaf ◽  
Mengjie Zhang ◽  
Bing Xue
Keyword(s):  
Machine Learning ◽  
Image Analysis ◽  
Genetic Programming ◽  
Image Classification ◽  
Search Space ◽  
Machine Learning Techniques ◽  
Efficient Manner ◽  
Speeded Up Robust Features ◽  
Learning Techniques ◽  
High Level

© Springer International Publishing Switzerland 2016. Image analysis is a key area in the computer vision domain that has many applications. Genetic Programming (GP) has been successfully applied to this area extensively, with promising results. Highlevel features extracted from methods such as Speeded Up Robust Features (SURF) and Histogram of Oriented Gradients (HoG) are commonly used for object detection with machine learning techniques. However, GP techniques are not often used with these methods, despite being applied extensively to image analysis problems. Combining the training process of GP with the powerful features extracted by SURF or HoG has the potential to improve the performance by generating high-level, domain-tailored features. This paper proposes a new GP method that automatically detects different regions of an image, extracts HoG features from those regions, and simultaneously evolves a classifier for image classification. By extending an existing GP region selection approach to incorporate the HoG algorithm, we present a novel way of using high-level features with GP for image classification. The ability of GP to explore a large search space in an efficient manner allows all stages of the new method to be optimised simultaneously, unlike in existing approaches. The new approach is applied across a range of datasets, with promising results when compared to a variety of well-known machine learning techniques. Some high-performing GP individuals are analysed to give insight into how GP can effectively be used with high-level features for image classification.

scholarly journals Grammar Filtering for Syntax-Guided Synthesis

2020 ◽  
Vol 34 (02) ◽  
pp. 1611-1618
Author(s):  
Kairo Morton ◽  
William Hallahan ◽  
Elven Shum ◽  
Ruzica Piskac ◽  
Mark Santolucito
Keyword(s):  
Machine Learning ◽  
Programming Languages ◽  
Automated Reasoning ◽  
Search Space ◽  
Machine Learning Techniques ◽  
Learning Approaches ◽  
Domain Specific ◽  
Learning Techniques ◽  
Data Driven Approach ◽  
Domain Specific Knowledge

Programming-by-example (PBE) is a synthesis paradigm that allows users to generate functions by simply providing input-output examples. While a promising interaction paradigm, synthesis is still too slow for realtime interaction and more widespread adoption. Existing approaches to PBE synthesis have used automated reasoning tools, such as SMT solvers, as well as works applying machine learning techniques. At its core, the automated reasoning approach relies on highly domain specific knowledge of programming languages. On the other hand, the machine learning approaches utilize the fact that when working with program code, it is possible to generate arbitrarily large training datasets. In this work, we propose a system for using machine learning in tandem with automated reasoning techniques to solve Syntax Guided Synthesis (SyGuS) style PBE problems. By preprocessing SyGuS PBE problems with a neural network, we can use a data driven approach to reduce the size of the search space, then allow automated reasoning-based solvers to more quickly find a solution analytically. Our system is able to run atop existing SyGuS PBE synthesis tools, decreasing the runtime of the winner of the 2019 SyGuS Competition for the PBE Strings track by 47.65% to outperform all of the competing tools.

scholarly journals Genetic programming for region detection, feature extraction, feature construction and classification in image data

2020 ◽  
Author(s):  
Andrew Lensen ◽  
Harith Al-Sahaf ◽  
Mengjie Zhang ◽  
Bing Xue
Keyword(s):  
Machine Learning ◽  
Image Analysis ◽  
Genetic Programming ◽  
Image Classification ◽  
Search Space ◽  
Machine Learning Techniques ◽  
Efficient Manner ◽  
Speeded Up Robust Features ◽  
Learning Techniques ◽  
High Level

© Springer International Publishing Switzerland 2016. Image analysis is a key area in the computer vision domain that has many applications. Genetic Programming (GP) has been successfully applied to this area extensively, with promising results. Highlevel features extracted from methods such as Speeded Up Robust Features (SURF) and Histogram of Oriented Gradients (HoG) are commonly used for object detection with machine learning techniques. However, GP techniques are not often used with these methods, despite being applied extensively to image analysis problems. Combining the training process of GP with the powerful features extracted by SURF or HoG has the potential to improve the performance by generating high-level, domain-tailored features. This paper proposes a new GP method that automatically detects different regions of an image, extracts HoG features from those regions, and simultaneously evolves a classifier for image classification. By extending an existing GP region selection approach to incorporate the HoG algorithm, we present a novel way of using high-level features with GP for image classification. The ability of GP to explore a large search space in an efficient manner allows all stages of the new method to be optimised simultaneously, unlike in existing approaches. The new approach is applied across a range of datasets, with promising results when compared to a variety of well-known machine learning techniques. Some high-performing GP individuals are analysed to give insight into how GP can effectively be used with high-level features for image classification.

scholarly journals A Framework for Long-Term Learning Systems

2017 ◽  
Author(s):  
Diana Benavides-Prado
Keyword(s):  
Machine Learning ◽  
Computer Science ◽  
Efficient Algorithms ◽  
Learning Systems ◽  
Machine Learning Techniques ◽  
Intelligent Machines ◽  
Learning Techniques ◽  
The World ◽  
Design And Application

Increasing amounts of data have made the use of machine learning techniques much more widespread. A lot of research in machine learning has been dedicated to the design and application of effective and efficient algorithms to explain or predict facts. The development of intelligent machines that can learn over extended periods of time, and that improve their abilities as they execute more tasks, is still a pending contribution from computer science to the world. This weakness has been recognised for some decades, and an interest to solve it seems to be increasing, as demonstrated by recent leading work and broader discussions at main events in the field [Chen and Liu, 2015; Chen et al., 2016]. Our research is intended to help fill that gap.

Design synthesis knowledge and inductive machine learning

2001 ◽  
Vol 15 (3) ◽  
pp. 233-249 ◽  
Author(s):  
S. POTTER ◽  
M.J. DARLINGTON ◽  
S.J. CULLEY ◽  
P.K. CHAWDHRY
Keyword(s):  
Machine Learning ◽  
Conceptual Design ◽  
Design Process ◽  
Learning Algorithm ◽  
Early Stage ◽  
Machine Learning Techniques ◽  
Design Synthesis ◽  
Conceptual Design Phase ◽  
Learning Techniques ◽  
Inductive Machine Learning

A crucial early stage in the engineering design process is the conceptual design phase, during which an initial solution design is generated. The quality of this initial design has a great bearing on the quality and success of the produced artefact. Typically, the knowledge required to perform this task is only acquired through many years of experience, and so is often at a premium. This has led to a number of attempts to automate this phase using intelligent computer systems. However, the knowledge of how to generate designs has proved difficult to acquire directly from human experts, and as a result, is often unsatisfactory in these systems. The application of inductive machine learning techniques to the acquisition of this sort of knowledge has been advocated as one approach to overcoming the difficulties surrounding its capture. Rather than acquiring the knowledge from human experts, the knowledge would be inferred automatically from a set of examples of the design process. This paper describes the authors' investigations into the general viability of this approach in the context of one particular conceptual design task, that of the design of fluid power circuits. The analysis of a series of experiments highlights a number of issues that would seem to arise regardless of the working domain or particular machine learning algorithm used. These issues, presented and discussed here, cast serious doubts upon the practicality of such an approach to knowledge acquisition, given the current state of the art.

Sign in / Sign up

Export Citation Format

Share Document