37,949 Hits in 5.9 sec

Fast Automatic Heuristic Construction Using Active Learning [chapter]

William F. Ogilvie, Pavlos Petoumenos, Zheng Wang, Hugh Leather
2015 Lecture Notes in Computer Science  
Our approach, on the other hand, uses active learning to select and only focus on the most useful training examples.  ...  In this work, we present a low-cost predictive modelling approach for automatic heuristic construction which significantly reduces this training overhead.  ...  We demonstrate the effectiveness of our approach by using active learning to automatically construct a heuristic to determine which processor will give the better performance on a Cpu-Gpu based heterogeneous  ... 
doi:10.1007/978-3-319-17473-0_10 fatcat:gwcklt44kvcgfhgpirsh76lvne

On dual processing and heuristic approaches to moral cognition

Daniel K. Lapsley, Patrick L. Hill
2008 Journal of Moral Education  
), fast-and-frugal moral heuristics (Gigerenzer), schema accessibility (Lapsley & Narvaez) and moral expertise (Narvaez).  ...  We examine the implications of dual-processing theories of cognition for the moral domain, with particular emphasis upon 'System 1' theories: the Social Intuitionist Model (Haidt), moral heuristics (Sunstein  ...  For example, activation of a social construct (e.g.  ... 
doi:10.1080/03057240802227486 fatcat:v2wiu4xgvneorlew6c4fyyfwze

End-to-End Deep Learning of Optimization Heuristics

Chris Cummins, Pavlos Petoumenos, Zheng Wang, Hugh Leather
2017 2017 26th International Conference on Parallel Architectures and Compilation Techniques (PACT)  
Machine learning is a proven technique for learning such heuristics, but its success is bound by the quality of the features used.  ...  Our work introduces a better way for building heuristics. We develop a deep neural network that learns heuristics over raw code, entirely without using code features.  ...  To make heuristic construction fast and cheap, we must take humans out of the loop.  ... 
doi:10.1109/pact.2017.24 dblp:conf/IEEEpact/CumminsP0L17 fatcat:p6sckrwd6rcxbpuqh64gcow3li

Towards learning domain-independent planning heuristics [article]

Pawel Gomoluch, Dalal Alrajeh, Alessandra Russo, Antonio Bucchiarone
2017 arXiv   pre-print
Heuristic approaches are necessary to solve all but the simplest problems. In this work, we explore the possibility of obtaining domain-independent heuristic functions using machine learning.  ...  The challenge is therefore the learning of (corrections of) domain-independent heuristics that can be reused across different planning domains.  ...  The learning process leverages features computed based on graphs constructed by the Fast Forward (FF) heuristic [Hoffmann and Nebel, 2001] .  ... 
arXiv:1707.06895v1 fatcat:s2ftzqzxungenhpdu2hnecpg3m

Generative hyper-heuristics

Daniel R. Tauritz, John Woodward
2022 Proceedings of the Genetic and Evolutionary Computation Conference Companion  
Train and test on problem instances drawn from the same probability distribution (like machine learning). Constructing an optimizer is machine learning (this approach prevents "cheating").  ...  Optimizing Multi-level Partitioning • Extract functionality from existing graph partitioning techniques • Use Genetic Programming (GP) to construct new algorithms Hyper-heuristic Approach John R.  ... 
doi:10.1145/3520304.3533646 fatcat:7s3q45ysfjg5bigcf5aolgnlya

A vignette model for distributed teaching and learning

Marcel Chaloupka, Tony Koppi
1998 Research in Learning Technology  
Vignettes are small, first-principle, first-person, heuristic activities (which are mimetic) from which courses can be constructed Vignettes use an objectorientated approach to the development of computer-based  ...  learning materials.  ...  Vignettes can be defined as small, first-principle, first-person, heuristic activities (components), from which courses are constructed.  ... 
doi:10.1080/0968776980060107 fatcat:k44w2rsgsrc2tanvfucras7zhi

Pre-Clerkship Medical Students' Experiences and Perspectives of System 1 and System 2 Thinking: A Qualitative Study

William Sanders, Douglas McHugh
2021 Education Sciences  
heuristics, and prior experience of attaining competence.  ...  Six major themes emerged from analysis of participants' interview transcripts: cognitive processes, deliberate practice, learning environment: being alone or being together, stickiness factor, biases and  ...  (S02) Another student leveraged ANKI (; developer: Damian Elmes), software that supports spaced repetition learning, to pursue fast, automatic recall: Learning activities that brought  ... 
doi:10.3390/educsci11020034 fatcat:wl6zpihakzh7plvsvqaxhc4swy

Toward Intelligent Network Optimization in Wireless Networking: An Auto-learning Framework [article]

Wenyu Zhang, Zhenjiang Zhang, Han-Chieh Chao, Mohsen Guizani
2018 arXiv   pre-print
As such, in this article we propose an auto-learning framework (ALF) to achieve intelligent and automatic network optimization by using machine learning (ML) techniques.  ...  We review the basic concepts of ML techniques, and propose their rudimentary employment models in WCSs, including automatic model construction, experience replay, efficient trial-and-error, RL-driven gaming  ...  We propose to use supervised learning techniques to automatically conduct the model constructions process in NOPs.  ... 
arXiv:1812.08198v1 fatcat:tin2avfzrzbybb4aepgpvmvmcu

A vignette model for distributed teaching and learning

Marcel Chaloupka, Tony Koppi
2011 Research in Learning Technology  
One evident area of convergence in education has been the use of the Internet.  ...  This could imply that the full potential of distributed learning through convergence might never be achieved.  ...  Vignettes can be defined as small, first-principle, first-person, heuristic activities (components), from which courses are constructed.  ... 
doi:10.3402/rlt.v6i1.10983 fatcat:ixjqjmlaxfdlhld62dt2cxyvya

Fast and Slow Processes Underlying Theories of Information Technology Use

Thomas Ferratt, Jayesh Prasad, E. James Dunne
2018 Journal of the AIS  
Consistent with that conclusion, our EUP posits that, as learning from prior use occurs, the influence of fast, automatic, unconscious (type 1) cognitive processes increases while the influence decreases  ...  The intervention potential is highest for initial use of the target IT and lowest when learning is high such that use of the IT has become automatic.  ...  These variations on previous system use have implications for learning and the influence of underlying fast and slow processes. We discuss each construct below.  ... 
doi:10.17705/1jais.00482 fatcat:onm5sgq4m5gdxgt6o3jqxdejma

Artificial Intelligence at MITRE

Joseph L. Katz
1985 The AI Magazine  
Knowledge-Based Query Optimization The object of this project is to investigate the applicability of knowledge-based automatic programming technology to generating fast relational database queries using  ...  The system is derived from the architecture used in KNOBS, which was constructed to aid air mission planning.  ... 
dblp:journals/aim/Katz85 fatcat:olxai7lol5fkdoyh63xdlzvwka

A Connectionist Model of Attitude Formation and Change

Frank Van Overwalle, Frank Siebler
2005 Personality and Social Psychology Review  
We use an autoassociative network architecture with a linear activation update and the delta learning algorithm for adjusting the connection weights.  ...  The network is applied to well-known experiments involving deliberative attitude formation, as well as the use of heuristics of length, consensus, expertise, and mood.  ...  Let us first elaborate on the prior heuristic learning phase.  ... 
doi:10.1207/s15327957pspr0903_3 pmid:16083362 fatcat:z2yob7yfkncnnmm4d3b5thecoq

Introduction to automated design of scheduling heuristics with genetic programming

Marko Ðurasević, Domagoj Jakobović, Yi Mei, Su Nguyen, Mengjie Zhang
2022 Proceedings of the Genetic and Evolutionary Computation Conference Companion  
His research interests include the field of evolutionary computing, optimization methods, machine learning, and scheduling problems. He has published nineteen journal and conference papers.  ...  Domagoj Jakobović received his PhD degree in 2005 at the Faculty of Electrical Engineering and Computing, University of Zagreb, on the subject of generating scheduling heuristics with genetic programming  ...  Similar to machine learning, using at least two data sets Training set: Used during evolution to train dispatching rules Needs to be general enough Different ways of constructing it: using same instances  ... 
doi:10.1145/3520304.3533667 fatcat:nnzdi5llqjdg5go25bkwxygi5m

Improving heuristic mini-max search by supervised learning

Michael Buro
2002 Artificial Intelligence  
This approach allows an automatic, data driven exploration of the feature space.  ...  Third, an opening book framework is presented that enables programs to improve upon previous play and to explore new opening lines by constructing and searching a game-tree based on evaluations of played  ...  This is somewhat surprising, because already in the 1950s Samuel proposed a way for automatically tuning weights [21] similar to TD-learning.  ... 
doi:10.1016/s0004-3702(01)00093-5 fatcat:vzfhumuzyngrjihjr2mdbdhdsm

Assessing creative problem-solving with automated text grading

Hao-Chuan Wang, Chun-Yen Chang, Tsai-Yen Li
2008 Computers & Education  
The reliable performance in automatic concept identification and numeric grading demonstrates the potential of using automated grading to support the use of open-ended questions in science assessments  ...  language responses automatically.  ...  Rosé, Yuen-Hsien Tseng, and Chao-Lin Liu for their advices on techniques of language processing and machine learning.  ... 
doi:10.1016/j.compedu.2008.01.006 fatcat:bvo43pmetrgbhfvciwar4pcnjy
« Previous Showing results 1 — 15 out of 37,949 results