Ladder Networks: Learning under Massive Label Deficit
Journal Title: International Journal of Advanced Computer Science & Applications - Year 2017, Vol 8, Issue 7
Abstract
Advancement in deep unsupervised learning are finally bringing machine learning close to natural learning, which happens with as few as one labeled instance. Ladder Networks are the newest deep learning architecture that proposes semi-supervised learning at scale. This work discusses how the ladder network model successfully combines supervised and unsupervised learning taking it beyond the pre-training realm. The model learns from the structure, rather than the labels alone transforming it from a label learner to a structural observer. We extend the previously-reported results by lowering the number of labels, and report an error of 1.27 on 40 labels only, on the MNIST dataset that in a fully supervised setting, uses 60000 labeled training instances.
Authors and Affiliations
Behroz Mirza, Tahir Syed, Jamshed Memon, Yameen Malik
Keywords
Related Articles
- EP ID EP260686
- DOI 10.14569/IJACSA.2017.080769
- Views 86
- Downloads 0
Ranking Method in Group Decision Support to Determine the Regional Prioritized Areas and Leading Sectors using Garrett Score
The main objective of regional development is to achieve equal development in different regions. However, the long duration and complexity of the process may result in the unequal development of some regions. In order to...
Efficient Smart Emergency Response System for Fire Hazards using IoT
The Internet of Things pertains to connecting currently unconnected things and people. It is the new era in transforming the existed systems to amend the cost effective quality of services for the society. To support Sma...
Implementing and Comparison between Two Algorithms to Make a Decision in a Wireless Sensors Network
The clinical presentation of acute CO poisoning and hydrocarbon gas (Butane CAS 106-97-8) varies depending on terrain, humidity, temperature, duration of exposure and the concentration of gas toxic: From then consciousne...
A Minimum Redundancy Maximum Relevance-Based Approach for Multivariate Causality Analysis
Causal analysis, a form of root cause analysis, has been applied to explore causes rather than indications so that the methodology is applicable to identify direct influences of variables. This study focuses on observati...
Optimization based Approach for Content Distribution in Hybrid Mobile Social Networks
This paper presents the new strategy for smooth content distribution in the mobile social network. We proposed a new method, hybrid mobile social network architecture scheme considered as one node of a social community c...