Prabhanjan Mutalik

• Master of Science
• Research Scientist at KTH Royal Institute of Technology

Time-series prediction is important in diverse fields. Traditionally, methods for time-series prediction were based on fixed linear models because of mathematical tractability. Researchers turned their attention to artificial neural networks due to their better approximation capability. In this paper, we use feedforward neural networks with a singl...

The Hippocampus is a brain region responsible for learning, memory and spatial navigation. In that, the interactions between the CA3 and CA1 subregions have been the most studied due to the interesting dynamics between the two regions. The excitatory auto-associative connections in the CA3 and the lack thereof in CA1 can be modelled as an Echo Stat...

Animals exhibit sequential behavior in various forms such as grooming, navigation, planning etc. The sequential nature of animal behavior suggests that neuronal activity underlying such behaviors should also be organized in a sequential manner. Indeed, sequences of neuronal activity are a ubiquitous phenomenon across the brain regions across specie...

Malaria is an incredibly complex and relevant problem. It has afflicted humanity since centuries: disintegrating civilisations, decimating the population, causing political and demographic shifts coupled widespread desolation and poverty. With the advent of 20th century, the disease mechanism became more apparent leading to significant control acti...

The aim of the project was to measure impact of language similarity on the machine translation. We considered the perplexity value of English-German and English-French translations as a metric of similarity. LSTM based Sequence-to-Sequence model with attention mechanism was used to perform the translation on the WMT’15 Europarl corpus. The underlyi...

Discussion of ideas and issues related to Connectomics.

Neural Networks have been of significant interest to Machine Learning (ML) scientists and Computational Neuroscientists alike. But the way they have been used by these two schools of thought is different, with neuroscientists opting for biological fidelity and the computer scientists preferring computational efficiency over everything else. This ha...

Predicting the future course of a sequential collection of an observable has several applications in diverse fields. Traditional techniques assume fixed linear models. In contrast, models based on artificial neural networks are adaptive and nonlinear; however these are typically trained offline. This paper focuses on time-series prediction using a...

Predicting the future of a time-series is important in several and diverse applications. Whilst traditional methods are based on fixed linear models, techniques that use artificial neural networks are typically trained offline. We propose an online technique for time-series prediction using a single-layer feedforward neural network trained as an ex...