Machine Learning, Optimization, and Big Data

Name: Machine Learning, Optimization, and Big Data | Third International Conference, MOD 2017, Volterra, Italy, September 14-17, 2017, Revised Selected Papers
Brand: Springer
Price: 53.49 EUR
Availability: OnlineOnly

Third International Conference, MOD 2017, Volterra, Italy, September 14-17, 2017, Revised Selected Papers

Giuseppe Nicosia Panos Pardalos Giovanni Giuffrida Renato Umeton(Editor)

Springer (Publisher)

Published on 19. December 2017

XXI, 600 pages

E-Book

PDF with digital watermarking

System requirements

978-3-319-72926-8 (ISBN)

€53.49incl. 7% vat

System requirements

for PDF with digital watermarking

E-Book Single Licence

Available for download

Description

More details

Other editions

Content

Intro
Preface
Organization
Contents
Recipes for Translating Big Data Machine Reading to Executable Cellular Signaling Models
Abstract
1 Introduction
2 Background
2.1 Cellular Networks
2.2 Modeling Approach
2.3 Framework Overview
3 Model Representation Format
4 From Reading to Model
4.1 Simple Interaction Translation
4.2 Translation of Translocation Interaction
4.3 Translation of Complexes
4.4 Translation of Nested Interactions
4.5 Translation of Direct and Indirect Interactions
4.6 Translation from Table Reading Output
5 Matching Reading and Modeling
5.1 Protein Families
5.2 Cell Type
5.3 Cellular Location
5.4 Contradicting Interaction Type
5.5 Negative Information
6 Case Study
7 Conclusion
References
Improving Support Vector Machines Performance Using Local Search
1 Introduction
2 Support Vector Machines
3 Iterated Local Search
4 Our ILS Method for SVM Parameters Tuning
5 Experimental Analysis
6 Conclusions and Future Research
References
Projective Approximation Based Quasi-Newton Methods
1 Introduction
2 Preliminaries
2.1 Notation Remarks
2.2 Quadratic Response Surface Methodology
2.3 Quasi-Newton Optimization Methods
3 Algorithm Descriptions
4 Theoretical Ground
5 Modelling
6 Conclusion
A Proofs
References
Intra-feature Random Forest Clustering
Abstract
1 Introduction
2 The Algorithm
3 Performance Evaluation
4 Conclusions
References
Dolphin Pod Optimization
1 Introduction
2 Dolphin Pod Optimization
3 DPO Setting Parameters
4 Performance Metrics
5 Numerical Results
5.1 Analytical Benchmark Functions
5.2 Hull-Form SBD Optimization Problem
6 Conclusions and Future Work
References
Contraction Clustering (RASTER)
1 Introduction
2 Problem Description
2.1 The Clustering Problem
2.2 Motivating Use Case
2.3 Limitations of Common Clustering Methods
3 RASTER
3.1 High-Level Description
3.2 Tiles and RASTER Clusters
3.3 The Algorithm
3.4 Parallel RASTER
3.5 Generalizing to Higher Dimensions
3.6 Minimum Cluster Size in Disadvantageous Grid Layouts
4 Results
4.1 Ideal Data
4.2 Sample Datasets
4.3 Empirical Runtime
5 Related Work
6 Future Work
References
Deep Statistical Comparison Applied on Quality Indicators to Compare Multi-objective Stochastic Optimization Algorithms
1 Introduction
2 Related Work
3 Deep Statistical Comparison
4 Results and Discussion
4.1 Experimental Setup
4.2 First Experiment
4.3 Second Experiment
5 Conclusion
References
On the Explicit Use of Enzyme-Substrate Reactions in Metabolic Pathway Analysis
1 Introduction
1.1 A Nash Equilibrium Approach to Metabolic Pathways
1.2 Element Mass Balances and Charge Balancing
2 Explicitly Incorporating Enzyme-Substrate Reactions
2.1 Enzyme-Substrate Reactions
2.2 An Example of Binding and Unbinding Reactions
2.3 Multiple Minima from Protein Docking
2.4 A Multi-scale Methodology for Including Enzyme-Substrate Reactions
2.5 Enzyme Activity
3 Numerical Results
4 Conclusions
References
A Comparative Study on Term Weighting Schemes for Text Classification
1 Introduction
2 Text Classification
3 Classifiers
4 Results and Discussion
4.1 Experiments
4.2 Evaluation
4.3 Results
5 Conclusion
References
Dual Convergence Estimates for a Family of Greedy Algorithms in Banach Spaces
1 Introduction
2 Greedy Algorithms
3 Primal Convergence Results
4 Duality Gap and Convergence Result
5 Conclusion
References
Nonlinear Methods for Design-Space Dimensionality Reduction in Shape Optimization
1 Introduction
2 Dimensionality Reduction Methods
2.1 General Definitions and Assumptions
2.2 Principal Component Analysis
2.3 Kernel Principal Component Analysis
2.4 Local Principal Component Analysis
2.5 Deep Autoencoders
3 Shape Modification of a Destroyer Hull
4 Numerical Results
4.1 Evaluation Metrics
4.2 Evaluation of Design-Space Dimensionality Reduction Capabilities
5 Conclusions and Future Work
References
A Differential Evolution Algorithm to Develop Strategies for the Iterated Prisoner's Dilemma
1 Introduction
2 Differential Evolution: A Short Overview
3 Prisoner's Dilemma
3.1 Iterated PD and Benchmark Strategies
4 DE Develops IPD Strategies
4.1 The DE Approach with Memory
5 IPD Experiments
6 Conclusions
References
Automatic Creation of a Large and Polished Training Set for Sentiment Analysis on Twitter
1 Introduction
2 Related Work
3 Methodology
3.1 Training Set Creation
3.2 Classification
3.3 Dataset Pruning
4 Results
4.1 Test Set
4.2 Accuracy
5 Conclusion
References
Forecasting Natural Gas Flows in Large Networks
1 Introduction
1.1 Literature Review
1.2 The Data Set
1.3 Input Features
1.4 The Network
1.5 Evaluation
2 Recurrent Neural Network (RNN) with Design of Experiments (DOE) and Simulated Annealing
2.1 The Experiment
2.2 Optimal Design with Simulated Annealing
3 Recurrent Neural Network (RNN) with Genetic Algorithm (GA)
4 Conclusion
References
A Differential Evolution Algorithm to Semivectorial Bilevel Problems
Abstract
1 Introduction
2 The SVBLP: Optimistic vs. Pessimistic Approaches
3 Optimistic and Pessimistic Frontiers
4 A Differential Evolution Algorithm for the SVBLP
5 Computational Experiment
6 Conclusions
Acknowledgment
References
Evolving Training Sets for Improved Transfer Learning in Brain Computer Interfaces
1 Introduction
2 Related Work on Transfer Learning in BCI
2.1 Ensembles
2.2 ELGI
3 Methodology
3.1 P300 Speller Paradigm
3.2 Dataset Recordings
3.3 Prefiltering
3.4 Classifier
3.5 Conditions
3.6 Compared Algorithms
4 Evolved ELGI Ensemble
5 Results
6 Discussion and Conclusion
References
Hybrid Global/Local Derivative-Free Multi-objective Optimization via Deterministic Particle Swarm with Local Linesearch
1 Introduction
2 Optimization Problem Formulation
3 Performance Metrics
4 Hybrid Global/Local Deterministic Algorithm
4.1 MODPSO
4.2 DFMO
4.3 MODHA
4.4 Algorithm Parameters and Setup
5 Numerical Results
5.1 Analytical Benchmark Problems
5.2 High-Speed Catamaran Optimization
6 Conclusions and Future Work
References
Artificial Bee Colony Optimization to Reallocate Personnel to Tasks Improving Workplace Safety
1 Introduction
2 Multi-objective Optimization
2.1 Non-dominated Sorting Bee Colony Optimization
3 Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS)
4 Worker's Risk Perception and Caution
5 Problem Formulation
5.1 Objectives
5.2 Problem Formulation
6 Experiments and Discussion
6.1 Dataset
6.2 Setup of the Parameters
6.3 Optimization Results
7 Conclusion
References
Multi-objective Genetic Algorithm for Interior Lighting Design
1 Introduction
2 The Inverse Lighting Problem
2.1 Blender as Direct Engine
3 Multi-objective Optimization
3.1 Previous Related Works
3.2 NSGA-II
3.3 Fitness Evaluation and Constraint Handling
4 The Proposed Strategy
5 Results
5.1 Art Gallery
5.2 Office
6 Conclusions
References
An Elementary Approach to the Problem of Column Selection in a Rectangular Matrix
1 Introduction
1.1 Historical Background
1.2 Our Contribution
2 Proof of Theorem 1.5
2.1 Suitable Choice of the Extracted Vectors
2.2 Controlling the Individual Eigenvalues
2.3 Controlling the Greatest Eigenvalue
2.4 Two Simple Examples
3 Computational Considerations
3.1 A Simple Algorithm
3.2 Scalability vs Accuracy
3.3 Extracting Representative Images from a Dataset
4 Conclusion
References
A Simple and Effective Lagrangian-Based Combinatorial Algorithm for S3VMs
1 Introduction and Related Work
1.1 The Semi-supervised Scenario
1.2 Continuous vs Combinatorial Approach
2 Lagrangian S3VM
2.1 Dealing with Hyper-parameters
2.2 Balance Constraint as a Guide
2.3 Inductive vs Transductive S3VMs
2.4 Method Details
3 Experiments
3.1 Algorithms
3.2 Datasets
3.3 Model Selection
3.4 Experimental Results
3.5 Technical Details
4 Conclusion and Remarks
References
A Heuristic Based on Fuzzy Inference Systems for Multiobjective IMRT Treatment Planning
Abstract
1 Introduction
2 Brief Review of the Literature
3 Multiobjective Optimization Problem
4 Heuristic Procedure Based on FIS
5 Illustration of the Application of the Procedure
6 Conclusions
Acknowledgments
References
Data-Driven Machine Learning Approach for Predicting Missing Values in Large Data Sets: A Comparison ...
Abstract
1 Introduction
2 Related Work
3 System Design
3.1 Data Source and Data Preparation
3.2 Methods for Imputation of Missing Values
4 Performance Measurements and Results
4.1 Algorithms Tuning
4.2 Evaluation Measures
4.3 Results and Considerations
5 Proposed Imputation Approach
6 Conclusions
References
Mineral: Multi-modal Network Representation Learning
1 Introduction
2 Preliminary
3 Mineral
4 Experiments and Results
4.1 Baselines
4.2 Link Prediction
4.3 Node Label Classification
4.4 Network Visualization
4.5 Parameter Sensitivity
5 Related Work
6 Conclusion
References
Visual Perception of Mixed Homogeneous Textures in Flying Pigeons
Abstract
1 Introduction
2 Background and Related Works
2.1 Visual Perception of Textures by Animals and Humans
2.2 Studying a Pigeon in Flight Using GPS and Brain Activity Loggers
2.3 The Recognition of Texture and Texture Elements in Motion
3 Materials and Methods
3.1 Spatial Datasets
3.2 Methods
3.3 QGIS Plugins
4 Results
4.1 Analysis of Typical Texture Frequencies
4.2 Comparison Texture Frequencies and Brain Activity in Flying Pigeon
5 Conclusions
References
Estimating Dynamics of Honeybee Population Densities with Machine Learning Algorithms
1 Introduction
1.1 State of the Art
1.2 Research Questions
2 Material and Methods
2.1 Hardware Description
2.2 Learning Algorithms
2.3 Dataset Description
2.4 Learning Experiments
3 Results
4 Discussion
5 Conclusions
6 Future Work
References
SQG-Differential Evolution for Difficult Optimization Problems under a Tight Function Evaluation Budget
Abstract
1 Introduction
2 Description of the SQG-DE Algorithm
2.1 Conceptual Description
2.2 Quantitative Description
3 A Comparison of Algorithm Performance on a Constrained Function Evaluation Budget
3.1 Algorithms and Test Functions
3.2 Performance Measures
3.3 Results Comparison
4 Discussion and Outlook
5 Conclusions
Acknowledgments
References
Age and Gender Classification of Tweets Using Convolutional Neural Networks
1 Introduction
2 Related Literature
2.1 PAN Editions
2.2 Word2vec
2.3 Previous System
2.4 Convolutional Neural Networks
3 Methodology
3.1 Pre-trained Vectors
3.2 Dataset
3.3 Preprocessing
3.4 Model
3.5 Model Variations
3.6 Evaluation
4 Results and Discussion
5 Conclusion and Recommendations
References
Approximate Dynamic Programming with Combined Policy Functions for Solving Multi-stage Nurse Rostering Problem
1 Introduction
2 The Multi-stage Nurse Rostering Problem
2.1 Problem Description
2.2 Problem Modification
3 Proposed Algorithm
3.1 Pre-process Phase
3.2 Local Phase - Value Function Approximation
3.3 Global Phase - Lookahead Policy
4 Experimental Design and Results Analysis
4.1 Experimental Settings
4.2 Lookahead Period Comparison
4.3 Algorithm Validation and Comparison
5 Conclusion
References
A Data Mining Tool for Water Uses Classification Based on Multiple Classifier Systems
Abstract
1 Introduction
2 Data and Study Area
2.1 United States Geological Survey (USGS)
2.2 Cauca River Modeling Project Phase II (PMC II)
2.3 Río Piedras Watershed (Test Dataset 1)
2.4 Illinois River (Test Dataset 2)
3 Models Description
3.1 Base Classifiers
3.2 MCS Models
4 Data Preprocessing
5 Results
6 Conclusions
Acknowledgements
References
Parallelized Preconditioned Model Building Algorithm for Matrix Factorization
1 Introduction
2 Preconditioned Model Building
2.1 A First Comparison with Other Optimizers
3 Parallelization of PMB-Based Matrix Factorization
3.1 Computational Tasks for Sparse Matrix Factorization
3.2 Storing the Sparse Matrix and Auxiliary Data in Memory
3.3 Efficient and Lock-Free Parallel Implementation of the Tasks
4 Experimental Results
5 Conclusions
References
A Quantitative Analysis on Required Network Bandwidth for Large-Scale Parallel Machine Learning
1 Introduction
2 Background
2.1 Parameter Exchange Methods for Large Scale Machine Learning Systems
2.2 SimGrid: A Distributed Environment Simulator
3 Network Model
4 Experiments
4.1 Structure of the Cluster and the Network
4.2 Parameter Exchange Methods
4.3 Results of Experiments
5 Discussion
5.1 Discussion on the Results
5.2 Data Size, Data Representation, Link Bandwidth
5.3 Computation Time and Parallelization Efficiency
6 Conclusion
References
Can Differential Evolution Be an Efficient Engine to Optimize Neural Networks?
1 Introduction
2 Background
2.1 Differential Evolution
3 Related Works
4 The Algorithm
5 Experimental Results
5.1 Datasets
5.2 Results
6 Conclusions and Future Works
References
BRKGA-VNS for Parallel-Batching Scheduling on a Single Machine with Step-Deteriorating Jobs and Release Times
Abstract
1 Introduction
2 Notations and Problem Description
3 The Problem {\varvec 1}\left| {{\varvec r}_{{\varvec j}} ,\
{\varvec p} - {\varvec batch},\
{\varvec p}_{{\varvec j}} = {\varvec a}_{{\varvec j}} \
{\varvec or}\
{\varvec a}_{{\varvec j}} {\,+\,} {\varvec b,}\
{\varvec c}} \right|{\varvec C}_{{{\varvec max}}}
3.1 The Properties of General Problem
3.2 The Properties of Special Cases
4 Heuristic
5 BRKGA-VNS Algorithm
5.1 Key Steps of BRKGA-VNS Algorithm
5.2 RVNS
5.3 Computational Experiments and Comparison
6 Conclusion
Acknowledgments
References
Petersen Graph is Uniformly Most-Reliable
1 Motivation
2 Uniformly Most-Reliable Graphs
3 Related Work
4 Petersen Graph
5 Conclusions and Trends for Future Work
References
GRASP Heuristics for a Generalized Capacitated Ring Tree Problem
1 Motivation
2 Capacitated Two-Node Survivable Tree Problem
3 GRASP Resolution
3.1 Construction Phase
3.2 Local Search Phase
4 Experimental Analysis
5 Conclusions and Trends for Future Work
References
Data-Driven Job Dispatching in HPC Systems
1 Introduction
2 Data Description
3 Job Duration Prediction
4 Job Dispatching Methods
5 Experimental Results
5.1 Prediction Performance
5.2 Dispatching Performance Using Prediction
6 Related Work
7 Conclusions
References
AbstractNet: A Generative Model for High Density Inputs
Abstract
1 Introduction
2 AbstractNet
2.1 The Model
2.2 Conditional AbstractNets
3 Experiments
3.1 Piano Dataset
3.2 Complex Samples
4 Conclusion
Acknowledgements
References
A Parallel Framework for Multi-Population Cultural Algorithm and Its Applications in TSP
1 Introduction
2 Preliminaries
3 Implementation of the Proposed Parallel Framework on GPUs
4 Experimental Results
5 Conclusion
References
Honey Yield Forecast Using Radial Basis Functions
1 Introduction
2 Honey Production and Weather Data in Andalusia
3 Radial Basis Functions Models
3.1 RBF Interpolation
3.2 Cross-Validation
4 Variable Screening
5 Computational Results
6 Conclusions
References
Graph Fragmentation Problem for Natural Disaster Management
1 Motivation
2 Background
3 Graph Fragmentation Problem
4 Analysis
5 Mathematical Programming Formulation
6 Bounds for the GFP
7 Proof of Concept
8 Conclusions and Trends for Future Work
References
Job Sequencing with One Common and Multiple Secondary Resources: A Problem Motivated from Particle Therapy for Cancer Treatment
1 Introduction
1.1 Contribution of This Work
2 Related Work
3 Problem Definition and Complexity
3.1 Computational Complexity
3.2 Lower and Upper Bounds
4 Least Lower Bound Heuristic
5 Mixed Integer Linear Programming Formulation
6 A* Algorithm
7 Computational Results
8 Conclusions
References
Robust Reinforcement Learning with a Stochastic Value Function
1 Introduction
2 Background
2.1 DDPG
2.2 TRPO
3 Proposal
4 Experiments
4.1 Experimental Methods
4.2 Sampling Efficiency
4.3 Robustness
5 Concluding Remarks
References
Finding Smooth Graphs with Small Independence Numbers
1 Introduction
2 Problem Formulation
3 Algorithmic Approach
3.1 Solution Representation
3.2 Core Algorithm
4 Bounds and Other Useful Properties
5 Checking Infeasibility
5.1 Symmetry Breaking
6 Computational Results
6.1 Problem 1
6.2 Problem 2
6.3 Problem 3
6.4 Problem 4
7 Conclusion and Further Work
References
BioHIPI: Biomedical Hadoop Image Processing Interface
1 Introduction
2 Apache Hadoop
2.1 Hadoop Distributed File System (HDFS)
2.2 MapReduce
2.3 Yet Another Resource Negotiator (YARN)
3 Hadoop Image Processing Interface (HIPI)
4 Biomedical Hadoop Image Processing Interface (BioHIPI)
4.1 BioHIPIImageBundle
4.2 BioHIPIImageHeader
4.3 BioHIPIImage
4.4 CodecManager
5 Experimental Analysis
6 Conclusion
References
Evaluating the Dispatching Policies for a Regional Network of Emergency Departments Exploiting Health Care Big Data
1 Introduction
2 Clinical Pathways and Health System Analysis
3 A DES Model Powered by the HCBD
3.1 The ED Network Operating in Piedmont Region
3.2 A Two-Phase DES Model
4 Quantitative Analysis
5 Conclusions and Future Developments
References
Refining Partial Invalidations for Indexed Algebraic Dynamic Programming
1 Introduction
2 Indexed Algebraic Dynamic Programming
2.1 Modelling a Simple Example
3 Partial Invalidation
4 Refining Partial Invalidation
5 Conclusion
References
Subject Recognition Using Wrist-Worn Triaxial Accelerometer Data
1 Introduction
2 Related Work
3 Wrist-Worn Triaxial Accelerometer Data
3.1 Dataset Description
3.2 Data Preparation
3.3 Feature Set
3.4 Training, Validation and Test Sets
3.5 Data Transformation
3.6 Class Labels and Performance Metrics
4 Experiments
4.1 Binary and Multi-class Classification Algorithms
4.2 One-Class Classification Algorithms
4.3 Feature Selection
4.4 Hyper-parameter Tuning
5 Results and Discussion
6 Conclusions and Further Work
References
Detection of Age-Related Changes in Networks of B Cells by Multivariate Time-Series Analysis
1 Introduction
2 Material and Methods
2.1 Dataset
2.2 Algorithm
3 Results
4 Conclusion and Ongoing Work
References
Author Index

System requirements

Save as PDF Copy link into clipboard

Schweitzer Fachinformationen

Machine Learning, Optimization, and Big Data

Description

More details

Other editions

Additional editions

Content

System requirements