AI2111 Smart Science AI-Powered Discoveries for Academic Minds 1
Investigate Global Trends in AI for Behavioral and Cognitive Sciences
Module 17 – Investigate global Trends
in AI for Behavioral and Cognitive
Sciences
At the end of this module, you are expected to:
1. Understand AI's Role in Shaping the Future of Behavioral and Cognitive
Sciences;
2. Implement AI for Mental Health Interventions; and
3. Evaluate AI's Role in Understanding Human Behavior:.
INTRODUCTION
Cognitive capacities are reflected in human behavior. Effective communication with others is
facilitated by the basic connection between human intellect and the many experiences or qualities of
consciousness/emotions, such as joy, sadness, wrath, etc. Understanding how to distinguish between ideas,
feelings, and behaviors is essential to developing emotional self-control and improving our ability to react
under pressure. Cognitive behavior is the capacity to perceive, analyze, process, interpret, recall, and
retrieve information while making decisions to react appropriately. Deception detection is one of the most
important areas to link human behavior, primarily in the forensic arena, after making a large impact on
emotion analysis. In advanced levels of behavioral science, detecting fraud, falsehoods, malevolent intent,
anomalous behavior, emotions, tension, etc., is important. The fields of pattern recognition, data extraction
and analysis, and interpretation have all benefited greatly from artificial intelligence and machine learning
(AI/ML).
COGNITIVE SCIENCE
To comprehend the cognitive functioning of the human mind and the underlying mental processes,
the field of study known as cognitive science interfaces various disciplines, including neuroscience,
computer science, psychology, artificial intelligence (AI), philosophy, linguistics, and anthropology.
Understanding the nature of the human mind and how it realizes, processes, transforms and manipulates
knowledge through mental representations is the main goal of this course.
To get a better understanding of how thought processes occur in the human brain, cognitive science
primarily depends on creating representative mental structures and examining the computing processes
that operate on them. In this case, mental models are comparable to computer data structures, and
computational processes are comparable to computational algorithms that work with the aforementioned
data structures. The field of cognitive science employs several methodologies to elucidate the
characteristics of mental models and computational processes. The most popular techniques consist of:
Course Module
AI2111 Smart Science AI-Powered Discoveries for Academic Minds 2
Investigate Global Trends in AI for Behavioral and Cognitive Sciences
• RULE-BASED MODELS
The rule-based method originated in the 1970s, in general. According to this paradigm, thinking is
the act of applying inference rules of the 'IF...THEN...' kind to symbols that represent sentence
patterns in language.
Consider the proverb, "You will get injured if you run too fast during a full marathon." When this
rule is applied to the sign "you ran too fast in a full marathon," the output symbol "you will suffer
from injuries" is produced. When another rule, "IF you suffer from injuries, THEN your body should
rest for 10 hours to recover," is applied to the latter, the output symbol "your body should rest for
10 hours to recover" is produced. When it comes to simulating the intricate aspects of human
thought processes like language usage and problem-solving, rule-based models are crucial. To
create expert systems that are useful, the medical industry has made substantial use of rule-based
techniques.
• CONNECTIONIST MODEL
Originally introduced in the 1980s, the connectionist approach focused on models of parallel-
distributed processing. These use artificial neural networks (ANN) to represent neural areas in the
brain to simulate human mental processes. Unlike rule-based approaches, connectionist models run
computer activities in parallel rather than serially.
Rule-based systems usually utilize inference rules to simulate mental events. Conversely,
connectionism is more focused on simultaneously satisfying a large number of criteria and data-
processing constraints. For this reason, the connectionist method is used in modern facial
recognition applications. Gaining insight into psychological ideas, like language learning, is essential
to comprehending human thought processes.
• THEORETICAL NEUROSCIENCE
When the rule-based model and the connectionism method were finally integrated, theoretical
neuroscience advanced to a new level. This occurred in the 1990s and 2000s, when brain-related
technologies and tools, like magnetic resonance imaging (MRI) and functional magnetic resonance
imaging (fMRI), which could monitor brain activity in real-time in experimental setups, were
developed, connecting cognitive science to neuroscience.
Furthermore, as theoretical neuroscience gained traction, computer models that could depict the
brain's firing neurons—or brain activity—were created. This assisted in deciphering the complex
process of decision-making, or high-level thinking, occurring in the brain's prefrontal cortex.
Furthermore, the amygdala area of the brain and human emotions were better-understood thanks
to these models.
Course Module
AI2111 Smart Science AI-Powered Discoveries for Academic Minds 3
Investigate Global Trends in AI for Behavioral and Cognitive Sciences
• BAYESIAN MODELS
To comprehend psychological characteristics including vision, motor control, social cognition,
learning, and others, cognitive science mostly depends on Bayesian models. The Bayes theorem,
which links probability variables to human comprehension, is the basis of these models. These
concepts have applicability in contemporary robotics as well.
• DEEP LEARNING MODELS
Since the 1950s, cognitive science has used artificial intelligence to decipher the workings of the
human mind. Furthermore, deep learning techniques are being employed in applications like object
identification, language translation, gaming, and other related fields thanks to recent developments
in AI.
Fundamentally, theoretical neuroscience and connectionist models serve as the cornerstones
around which deep learning is built. But it makes use of far more sophisticated and sophisticated
algorithms, as well as neural networks with several neural layers. Reinforcement learning is being
used to investigate deep learning, which goes one step further by combining learning with the
rewards from earlier instances to optimize the model.
One such reinforcement learning model is DeepMind's AlphaGo, which beat a human Go player in
2016.
Deep learning is currently influencing neuroscience and leading to important advancements in
cognitive psychology.
Aside from the previously described methods, cognitive theories of logical principles, ideas,
pictures, propositions, and analogies have been used to investigate mental representations and
computations that disclose the specifics of how the mind functions (i.e., how neurons fire when
executing certain tasks).
KEY COGNITIVE SCIENCE METHODS
To accomplish the scientific study of the mind and intellect, cognitive science unites several
perspectives and disciplines. Brain imaging technologies, computer modeling approaches, and human-
subject psychology investigations are the main methodologies used in cognitive science.
1. Psychological/behavioral experiments
Studying participants' behavioral reactions to different stimuli is essential to quantifying and
understanding intelligent behavior because it shows how the human brain processes inputs. This
can be accomplished by conducting participant-involved behavioral or psychological experiments.
These tests reveal the characteristics and preferences of users' behavior.
For the participants, these experiments assist in monitoring the following variables:
• Reaction time of the participant: This variable measure how long it takes for a user to respond when
a stimulus is shown. The temporal gap essentially reveals a person's capacity for cognition, their
mode of thought, and their response to stimuli—all of which speak to their core character.
Course Module
AI2111 Smart Science AI-Powered Discoveries for Academic Minds 4
Investigate Global Trends in AI for Behavioral and Cognitive Sciences
• Psychophysical response: In cognitive psychology, psychophysical experiments are studies in which
participants' reactions to certain stimuli are noted to gain a better understanding of their cognitive
and sensory biases. Experiments can be conducted, for instance, to ascertain the possible user
reaction to loud music or traffic horns. Additionally, it can detect how people react to noises that
are louder than a predetermined threshold or evaluate various color textures.
• Eye tracking: This variable is used to identify cognitive processes including language processing and
visual perception. A user's eyes move very little for a time when they are concentrated on anything.
Following the subject's gaze while they work on an online activity is one example. The observation
can provide light on how quickly decisions are made and provide an understanding of how the
individual in question thought through the options.
2. Brain Imaging
Brain imaging enables real-time observation of a subject's brain activity while they are carrying out
an activity. It provides insights into how the brain processes information by linking user behavior
to cognitive brain function. Techniques for brain imaging are essential to cognitive neuroscience.
3. Computational Modeling
A systematic description of an issue that makes sense to understand and solve is necessary for
computational modeling. This is where the previously described mental representation is useful.
Here, mental-like computational models are examined to assess behavioral or psychological tests
that are conducted. Once constructed, models are then utilized to measure people's cerebral
activity.
The aforementioned methodologies are employed as either independent computational models or
combined models to simulate cognitive abilities like speech, sight, hearing, etc. These models are essential
for improving our knowledge of human cognition, which helps us create intelligent systems with thinking
and decision-making capacities comparable to those of humans. Additional neurobiological techniques to
better comprehend the human mind include postmortem research, single-unit recording, and direct brain
stimulation. Through electrodes, a current is applied to a subject in direct brain stimulation to improve
cognitive function. Using a microelectrode setup, the single-unit recording approach records the electro-
physiological responses of a single neuron in the subject while the subject is doing a task.
Course Module
AI2111 Smart Science AI-Powered Discoveries for Academic Minds 5
Investigate Global Trends in AI for Behavioral and Cognitive Sciences
References and Supplementary Materials
1. Bhatt, P., Sethi, A., Tasgaonkar, V. et al. Machine learning for cognitive behavioral
analysis: datasets, methods, paradigms, and research directions. Brain Inf. 10, 18
(2023). [Link]
2. [Link]
3. [Link]
science/
Course Module