Lidia Ghosh - Cognitive Modeling of Human Memory and Learning

1. Chapter 2
2. Chapter 3
3. Chapter 4
4. Chapter 5
5. Chapter 6

1. Chapter 1
2. Chapter 2
3. Chapter 3
4. Chapter 4
5. Chapter 5
6. Chapter 6

IEEE Press

445 Hoes Lane

Piscataway, NJ 08854

IEEE Press Editorial Board

Ekram Hossain, Editor in Chief

Lidia Ghosh, Amit Konar, and Pratyusha Rakshit

Artificial Intelligence Laboratory
Department of Electronics and TeleCommunication Engineering
Jadavpur University, Kolkata700032, India

Copyright 2021 by The Institute of Electrical and Electronics Engineers, Inc. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey.

Published simultaneously in Canada.

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate percopy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 7508400, fax (978) 7504470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 7486011, fax (201) 7486008, or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 7622974, outside the United States at (317) 5723993 or fax (317) 5724002.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com.

Library of Congress CataloginginPublication Data

Names: Ghosh, Lidia, author. | Konar, Amit, author. | Rakshit, Pratyusha,

author.

Title: Cognitive modeling of human memory and learning : a noninvasive

brain-computer interfacing approach / Lidia Ghosh, Artificial

Intelligence Lab., Dept. of Electronics and TeleCommunication

Engineering, Amit Konar, Artificial Intelligence Lab., Dept. of

Electronics and TeleCommunication Engineering, Pratyusha Rakshit,

Artificial Intelligence Lab., Dept. of Electronics and

TeleCommunication Engineering.

Description: Hoboken, New Jersey : Wiley, [2021] | Includes bibliographical

references and index.

Identifiers: LCCN 2020015457 (print) | LCCN 2020015458 (ebook) | ISBN

9781119705864 (cloth) | ISBN 9781119705871 (adobe pdf) | ISBN

9781119705918 (epub)

Subjects: LCSH: Memory. | Braincomputer interfaces. | Cognitive

neuroscience.

Classification: LCC BF371 .G46 2021 (print) | LCC BF371 (ebook) | DDC

153.1/20113dc23

LC record available at https://lccn.loc.gov/2020015457

LC ebook record available at https://lccn.loc.gov/2020015458

Cover Design: Wiley

Cover Image: Paolo Carnassale/Getty Images

Existing works on human memory models take into account the behavioral perspectives of learning/memory and thus have limited scope in diagnostic and therapeutic applications of memory. The present title makes a humble attempt to model human memory from the cognitive standpoint by utilizing the brain activations acquired from the cortex by electroencephalography ( EEG ) and functional nearinfraredspectroscopy ( fNIRs ) means (during the course of subjective learning and memory recall). The EEGbased memory modeling is advantageous for its inherent merit to offer prompt temporal response (of memory) to perceptual cues. The prompt response of memory helps in understanding its correspondence with the stimulus, thereby justifying the selection of the EEG modality for the experimental protocol design for memory modeling. The fNIRs device, on the other hand, having good spatial resolution, is useful to accurately localize the regions of brain activations for selected memory tasks. Thus memory activation study with EEG preceded by localization of brain regions by fNIRs device for a given memory task is an ideal choice for memory modeling by experimental means. Although functional magnetic resonance imaging ( fMRI ) is a better choice for spatial localization of brain regions for memory tasks, in this book spatial localization is undertaken by fNIRs device only for its portability, userfriendliness, and costeffectiveness.

Like computer memory, human memory too maintains a hierarchy. For instance, information acquired by sense organs is temporarily stored into sensory registers for transfer into shortterm memory ( STM ) located in the prefrontal lobe. Next, depending on the relative importance of the information, sometimes the contents of STM are transferred into longterm memory ( LTM ), located in the hippocampus region. A third form of memory, referred to as working memory ( WM ), also resides in the prefrontal lobe. The WM is generally used to analyze information stored into STM for logical reasoning, information matching, and decision making. There exist signaling pathways from the STM to the WM to the LTM. There also exist direct routes from the STM to the LTM. These signaling pathways include long chain on neurons, forming deep brain networks. The book attempts to model the deep signaling pathways connecting the modules in the memory hierarchy by deep learning. The proposed models of memory developed with brain activations are advantageous in the early diagnosis and prognosis of certain brain diseases, such as the Alzheimer's disease, schizophrenia, prosopagnosia, and many others.

Computational intelligence ( CI ) is the rubric of a number of intelligent tools and techniques that synergistically complement each other's performance and thus jointly may serve as a complete approach to handle complex realworld problems. The memory encoding and recall processes and signal transduction across distributed modules of memory is primary controlled by the human nervous system. This can be taken up by artificial neural networks and deep learning models of CI. In addition, the brain signals being nonstationary have intra and intersession fluctuations, which can be modeled by fuzzy sets (in particular type2 fuzzy sets [ T2FS ]). In fact, the book proposes interesting models of memory and learning by amalgamating fuzziness in the settings of deep brain learning network. The parameter optimization of the memory models to attain the best performance can be designed by evolutionary computation. Thus memory modeling can be performed efficiently by synergism of different CI techniques.