Papers by Chandrasekhar Pammi
Bad to worse: Striatal coding of the relative value of painful decisions
Serial order processing or Sequence processing underlies many human activities such as speech, la... more Serial order processing or Sequence processing underlies many human activities such as speech, language, skill learning, planning, problem solving, etc. Investigating the neural bases of sequence processing enables us to understand serial order in cognition and helps us building intelligent devices. In the current paper, various cognitive issues related to sequence processing will be discussed with examples. Some of the issues are: distributed versus local representation, pre-wired versus adaptive origins of representation, implicit versus explicit learning, fixed/flat versus hierarchical organization, timing aspects, order information embedded in sequences, primacy versus recency in list learning and aspects of sequence perception such as recognition, recall and generation. Experimental results that give evidence for the involvement of various brain areas will be described. Finally, theoretical frameworks based on Markov models and Reinforcement Learning paradigm will be presented. These theoretical ideas are useful for studying sequential phenomena in a principled way.
In this paper, the applicability of using MultiLayer Perceptron Networks in symmetric block ciphe... more In this paper, the applicability of using MultiLayer Perceptron Networks in symmetric block ciphers is explored. A prototype symmetric block cipher is proposed. It employs a Multilayer-Perceptron (MLP) Network that decides on the algorithm used for encryption. The MLP Network is in turn dependent on the secret key. By employing a mutating algorithm comprising of cryptographically proven modular arithmetic and feistel networks, it is hoped that such a symmetric block cipher will be resistant to modern cryptanalytic attacks such as differential and linear attacks.
Neuroimage
Previous brain imaging studies investigating motor sequence complexity have mainly examined the e... more Previous brain imaging studies investigating motor sequence complexity have mainly examined the effect of increasing the length of pre-learned sequences. The novel contribution of this research is that we varied the structure of complex visuo-motor sequences along two different dimensions using mxn paradigm. The complexity of sequences is increased from 12 movements (organized as a 2 × 6 task) to 24 movements (organized as 4 × 6 and 2 × 12 tasks). Behavioral results indicate that although the success rate attained was similar across the two complex tasks (2 × 12 and 4 × 6), a greater decrease in response times was observed for the 2 × 12 compared to the 4 × 6 condition at an intermediate learning stage. This decrease is possibly related to successful chunking across sets in the 2 × 12 task. In line with this, we observed a selective activation of the fronto-parietal network. Shifts of activation were observed from the ventral to dorsal prefrontal, lateral to medial premotor and inferior to superior parietal cortex from the early to intermediate learning stage concomitant with an increase in hyperset length. We suggest that these selective activations and shifts in activity during complex sequence learning are possibly related to chunking of motor sequences.► Structure of complex motor sequences varied while controlling for sequence length. ► Chunking across several elements was observed with increase in long-range complexity. ► Concomitant shifts in fronto-parietal activation observed. ► Plausible neural correlates of chunking during motor sequence learning suggested.
Brain imaging data have so far revealed a wealth of information about neuronal circuits involved ... more Brain imaging data have so far revealed a wealth of information about neuronal circuits involved in higher mental functions like memory, attention, emotion, language etc. Our efforts are toward understanding the learning related effects in brain activity during the acquisition of visuo-motor sequential skills. The aim of this paper is to survey various methods and approaches of analysis that allow the characterization of learning related changes in fMRI data. Traditional imaging analysis using the Statistical Parametric Map (SPM) approach averages out temporal changes and presents overall differences between different stages of learning. We outline other potential approaches for revealing learning effects such as statistical time series analysis, modelling of haemodynamic response function and independent component analysis. We present example case studies from our visuo-motor sequence learning experiments to describe application of SPM and statistical time series analyses. Our review highlights that the problem of characterizing learning induced changes in fMRI data remains an interesting and challenging open research problem.
Bad to worse: Striatal coding of the relative value of painful decisions
Serial order processing or sequence processing underlies many human activities such as speech, la... more Serial order processing or sequence processing underlies many human activities such as speech, language, skill learning, planning, problem-solving, etc. Investigating the neural bases of sequence processing enables us to understand serial order in cognition and also helps in building intelligent devices. In this article, we review various cognitive issues related to sequence processing with examples. Experimental results that give evidence for the involvement of various brain areas will be described. Finally, a theoretical approach based on statistical models and reinforcement learning paradigm is presented. These theoretical ideas are useful for studying sequence learning in a principled way. This article also suggests a two-way process diagram integrating experimentation (cognitive neuroscience) and theory/ computational modelling (computational neuroscience). This integrated framework is useful not only in the present study of serial order, but also for understanding many cognitive processes.
Neuroimage, 2008
A decision maker may experience regret when a choice he makes results in a more adverse outcome t... more A decision maker may experience regret when a choice he makes results in a more adverse outcome than a different choice would have yielded. Analogously, he may experience rejoice when his choice resulted in better outcomes. We used fMRI to investigate the neural correlates of regret and rejoice where payoffs are in terms of a non-monetary medium. Incentives were created using painful outcomes in the form of mild electrical shocks to the foot and the possibility of avoiding them. We hypothesized that the neural response to a painful outcome resulting from an individual's choice would also reflect the degree of regret as measured by the likelihood that alternative choices would have yielded the same adverse outcome. Similarly, when an individual avoids a potential shock, he would experience a degree of rejoice that correlates with the probability he had of receiving the shock. For example, winning a bet when winning was unlikely, even if the outcome is the same, evokes more rejoice than winning when it was highly probable. Our results suggest that activation of a cortical network, consisting of the medial orbitofrontal cortex, left superior frontal cortex, right angular gyrus, and left thalamus, correlates with the degree of regret. A different network, including the rostral anterior cingulate, left hippocampus, left ventral striatum, and brainstem/midbrain correlated with rejoice. The right inferior orbitofrontal cortex, pre-supplementary motor area, anterior cingulate, and posterior cingulate showed similar patterns of activation with both regret and rejoice, suggesting that these regions may be associated with surprise from the realization of relatively unlikely events. Our results suggest that distinct, but overlapping networks are involved in the experiences of regret and rejoice.
Chunking Phenomenon in Complex Sequential Skill Learning in Humans
Sequential skill learning is central to much of human behaviour. It is known that sequences are h... more Sequential skill learning is central to much of human behaviour. It is known that sequences are hierarchically organized into several chunks of information that enables efficient performance of the acquired skill. We present clustering analysis on response times as subjects learn finger movement sequences of length 24 arranged in two ways – 12 sets of two movements each and 6 sets of four movements each. The experimental results and the analysis point out that greater amount of reorganization of sequences into chunks is more likely when the set-size is kept lower and discuss the cognitive implications of these findings.
It is well known that learning a sequential skill involves chaining a number of primitive actions... more It is well known that learning a sequential skill involves chaining a number of primitive actions together into chunks. We describe three different experiments using an explicit visuomotor sequence learning paradigm called the m×n task. The m × n task enables hierarchical learning of sequences by presenting m elements of the sequence at a time (called the set). The entire sequence to be learned is composed of n such sets and is called a hyperset. In the first experiment, we showed the chunking phenomenon while learning a sequence as opposed to following randomly generated visual cues. We further explored the nature of chunking across sets using complex sequences in the second experiment. Finally, we investigated effector dependence of the chunking patterns in the third experiment. Our results point out the facilitating factors for chunk formation in visuomotor sequence learning.
Uploads
Papers by Chandrasekhar Pammi