IJCNN 2015

144 papers
The 2015 International Joint Conference on Neural Networks (IJCNN) featured 173 papers, with a strong emphasis on the theoretical foundations of machine learning. Key research directions included active learning through version space reduction and convergence analysis, as well as the exploration of generalization bounds using sample compression and sparsity. Prolific authors contributing to these themes included José C. Príncipe, Badong Chen, and Michael Blumenstein. The year’s proceedings also highlighted diverse applications of neural architectures. Research spanned from computer vision, such as facial expression recognition using two-channel convolutional neural networks, to interactive reinforcement learning guided by speech. Other studies addressed autonomous navigation via neural path integration, the classification of migraine stages through EEG power, and system identification using kernel normalized mixed-norm algorithms. Practical implementations were also observed in traffic data analysis and HVAC thermal comfort control. This edition reflected a balance between advancing core learning theories and deploying neural models to solve complex engineering and medical problems.

Top Research Directions

Research directions most represented by IJCNN 2015 papers.

Preprocessing Sensitivity in Machine Learning Pipelines 1
Discretizing Continuous Features for Robust and Interpretable Models 1
Scale-Dependent Intrinsic Dimension 1
Hyperparameter Insensitivity in Regularized Learning 1
Overcoming Static Rule Rigidity in Dynamic Systems 1
Cost Aware Confidence Thresholding for Machine Learning Rejection 1
Closed Form Weight Optimization in Randomized Networks 1
Local Geometric Context for 3D Point Cloud Understanding 1
Expressivity Constraints in Grammar-Constrained Program Synthesis 1
Scalable Log Analysis for High Volume Distributed Systems 1
Tracking Structural Features with Persistent Homology 1
The Information Cost of Discretizing Continuous Data 1
Extracting Symbolic Rules from Neural Networks 1
Vectorizing Topological Signatures for Machine Learning 1
Moving Beyond Binary Classification Frameworks 1

Top Authors

AuthorPapers
Barbara Hammer5
Ammar Belatreche4
Badong Chen4
Furao Shen4
Michael Blumenstein4
Sergio Escalera4
Aurelio Uncini3
Hugo Jair Escalante3
Luca Oneto3
Nanning Zheng3
Stefan Wermter3
Umapada Pal3
André Luis Debiaso Rossi2
Artur Garcez2
Bernd Bischl2
Danilo Comminiello2
Dongbin Zhao2
Euntai Kim2
Giacomo Boracchi2
Guang-Bin Huang2

Impactful Papers

Papers from IJCNN 2015 that shaped research clusters.

Foundational

Influential

Papers

Title
New insights into the landscape relationships of host response to bacterial pathogens
On initial convergence behavior of the kernel least mean square algorithm
Online diffusion source detection in social networks
Online reinforcement learning by Bayesian inference
On the utility of sparse neural representations in adaptive behaving agents
Optimized deep belief networks on CUDA GPUs
PAC-Bayes Analysis for Twin Support Vector Machines
Pixel characteristics based feature extraction approach for roadside object detection
Probabilistic dynamic causal model for temporal data
Proposing a fast circular HOG descriptor for detecting rotated objects
Recognizing visual composite in real images
Recurrent convolutional neural networks for object-class segmentation of RGB-D video
Regularizing neural networks with adaptive local drop
Resource-constrained classification using a cascade of neural network layers
Restoring high frequency spectral envelopes using neural networks for speech bandwidth extension
Reward-based online learning in non-stationary environments: Adapting a P300-speller with a "backspace" key
Robust semi-supervised nonnegative matrix factorization
Runtime detection of activated polychronous neuronal group towards its spatiotemporal analysis
Scalable energy-efficient, low-latency implementations of trained spiking Deep Belief Networks on SpiNNaker
Self-organization of hippocampal representations in large environments
Self-structured confabulation network for fast anomaly detection and reasoning
Short answer question examination using an automatic off-line handwriting recognition system and a novel combined feature
Shrinkage learning to improve SVM with hints
Solving the data imbalance problem of P300 detection via Random Under-Sampling Bagging SVMs
Sparse density estimation on multinomial manifold combining local component analysis
Sparse uncorrelated cross-domain feature extraction for signal classification in brain-computer interfaces
Sparsity analysis of learned factors in Multilayer NMF
Spin-Transfer Torque Magnetic neuron for low power neuromorphic computing
Splitting with confidence in decision trees with application to stream mining
Stationarity of Matrix Relevance LVQ
Stochastic least squares learning for deep architectures
Stock price prediction based on stock-specific and sub-industry-specific news articles
Support vector machines and strictly positive definite kernel: The regularization hyperparameter is more important than the kernel hyperparameters
Tactile sequence classification using joint kernel sparse coding
The adaptable buffer algorithm for high quantile estimation in non-stationary data streams
The MC-ELM: Learning an ELM-like network with minimum VC dimension
Thermal comfort control based on MEC algorithm for HVAC systems
To tune or not to tune: Recommending when to adjust SVM hyper-parameters via meta-learning
Towards robust flood forecasts using neural networks
Transient phenomena prediction using recurrent neural networks
Two-layer Mixture of Factor Analyzers with Joint Factor Loading
Uncorrelated transferable feature extraction for signal classification in brain-computer interfaces
Use of ensembles of Fourier spectra in capturing recurrent concepts in data streams
Variable length concentration based feature construction method for spam detection

« prev Page 3/3