ABSTRACT
Understanding how a machine learns is a pressing topic as machine learning becomes more complex enabled by more powerful computers. This paper presents a visualization of neural networks to make them trackable during the operation of learning for pattern recognition, as well as testing for patterns. Specifically, our implementation includes fully connected neural networks, convolutional neural networks, and networks with memories. This will help us understand the insight of neural networks for pattern recognition to ensure full human control of the machines and to eliminate public's concern of recent leap in AI and machine learning. The visualization also helps to measure and identify performance bottleneck for future improvement.
- D. Bassu, P.W. Jones, L. Ness, and D. Shallcross, "Product Formalisms for Measures on Spaces with Binary Tree Structures: Representation, Visualization, and Multiscale Noise", https://arxiv.org/pdf/1601.02946v2.pdf, 2016.Google Scholar
- D.Bassu, R.Izmailov, A.McIntosh, L.Ness, D.Shallcross, "Centralized Multi---Scale Singular Vector Decomposition for Feature Construction in LIDAR Image Classification Problems", Applied Imagery Pattern Recognition Workshop, 2012. Google ScholarDigital Library
- Y. Rossikova, J, J. Li and P. Morreale, "Intelligent Data Mining for Translator Correctness Prediction", CloudSecurity2016, Columbia university NYC, April 2016.Google Scholar
- C. Chen, J. Iglasias, X. Lin, J. J. Li, P. Morreale, and L. Ness, "Facebook Traffic Pattern Analytics", 2016. Google ScholarDigital Library
- Andrew L. Maas, Awni Y. Hannun, and Andrew Y. Ng, "Rectifier Nonlinearities Improve Neural Network Acoustic Models", Proceedings of the 30th International Conference on Machine Learning, Atlanta, Georgia, USA, 2013. JMLR: W&CP volume 28.Google Scholar
- Sandeep Chakravorty and Amrita Chakrabarti, "Hidden Layer Optimization of Neural Network using Computational Technique", Proceedings of the International Conference on Advances in Computing, Communication and Control (ICAC3'09), pp 559--561, Mumbai, India - January 23 - 24, 2009. Google ScholarDigital Library
- Djork-Arné Clevert, Thomas Unterthiner and Sepp Hochreiter, "Fast and Accurate Deep Network Learning By Exponential Linear Units (ELUS)", Proceedings of the International Conference on Learning Representations 2016, San Juan, Puerto Rico, May 2-4, 2016.Google Scholar
Index Terms
- Visualizing Neural Networks for Pattern Recognition
Recommendations
Multi-class pattern classification using neural networks
Multi-class pattern classification has many applications including text document classification, speech recognition, object recognition, etc. Multi-class pattern classification using neural networks is not a trivial extension from two-class neural ...
Artificial neural network-based image pattern recognition
ACM-SE 30: Proceedings of the 30th annual Southeast regional conferenceThis paper addresses the use of a multi-layer fully-connected perceptron neural network for implementing a pattern recognizer. The input of the neural network is a set of seven standardized invariant moments in both the training procedure and ...
Hierarchical Autoassociative Polynimial Network (HAP Net) for pattern recognition
Extensive research and evaluations have been conducted on neural networks to improve classification accuracy and training time. Many classical architectures of neural networks have been modified in several different ways for advancement in design. We ...
Comments