Good day, Dr. Muhamed Jishad T K. I'm Arjun. I've had a look through your background and your professional record. I'm interested in hearing more about your work and how you see yourself fitting in here with our faculty. Ready to start?

Yes.

Could you share a brief overview of your professional journey in academia?

OK, so I have completed my PhD recently from National Institute of Technology, Calicut. My thesis was based on, you know.

Signal processing and classification for EEG signals for the applications in non-invasive brain computer interface.

Even in engineering college called Women Jodi Engineering College which is located in Kerala, I worked there for 2-2 years and prior to that joining the I was working as a project engineer at Energy Management Center Johanna. I worked for two years and I did my Mtech from magnetic allocate.

Prior to joining joining PhD, I was working as an Assistant Professor.

Specialization mass industry, power and automation from the Department of Electrical Engineering. My BTEC is in installation Control engineering for Nscologist engineering.

I don't know. Currently I'm working as a Junior Research fellow at Nat Calicut in a Ministry of Health Science of Conduct project, which is a robotic project working on underwater waves.

Thank you for sharing your academic and professional journey, Professor. It's clear you have a strong foundation in electrical engineering and research. Let's proceed.

Professor, let's begin by evaluating your expertise in artificial intelligence and machine learning in healthcare or closely related fields. Could you elaborate on your research involving AI or ML applications specifically designed for healthcare challenges?

MMM, of course. And actually my PC thesis is for classification of electroencephalographic signals, which is a biomedical signals which is taken from by placing electrodes over the skirt. It's actually measuring the activities inside our brain. So I was working on that easy signals captured from our brain and the major objective was to classify these eg signals based on the. Intention of the imagination of the.

Uh, subject. So I was working on a motor imagery based EEG, uh, signal classification. Uh, so I used machine learning techniques, uh, first and evaluated the performance. Uh, so that performance was not that good. The classification accuracy was not that much expected. So I'm going to neural networks. So I worked on convolution neural networks and advanced signal processing techniques for classifying this not only means to EGCPS, uh, for example, I was trying to classify.

Four tasks. The subject was imagining, uh, his left hand movement, uh, easy was taken corresponding to that particular movement and then right hand movement, uh, leg movement and 10 minutes. So personally, we have 4, uh, lab classes of easy data. I was trying to classify the four classes and, uh, trying to predict what the QSO was, uh, intending to do. So that was my project. I was used, I used, uh, convolution.

And similar costing techniques like continuous failure transform, uh, such techniques are used for uh, this.

My project which is actually applicable main application is for rehabilitation of patients with neuromuscular disorders or some spinal cord injury like situations. So that was my question.

Understood, Professor. This sounds like an impactful application of artificial intelligence and machine learning. To clarify further: What specific challenges did you encounter when transitioning from traditional machine learning approaches to convolutional neural networks for your EEG classification task, particularly in terms of data preprocessing and model optimization?

Actually, uh, my data is, uh, easy signals, which is time series data, uh, with, uh, lower multiple electrodes. So I had multiple, uh, time series data for a particular class itself. So while using machine learning techniques, I have to find, uh, find out some features using some techniques and then using that features only had to proceed further for training that model.

And I had to do artifact removals like egg is very prone to artifacts like the muscle actions, eye blinking and some other line noises like artifacts will be the ECG or electrocardiogram signal may interfere to be easy. So artifact removal was necessary for machine learning techniques and we have to find out the features and select useful features and then we have to give the features for training. But when?

Transcending to convolution neural networks, that reprocessing part was changed to some something else like in conventional networks that data should be great like so I have time series data, so I have to convert the time series data into some. Represent that in a like a matrix. So I was using convolution on networks so that.

Continuous, continuous failure transform. So that gave me the pattern to track matrix format and there were multiple time series better from different channels. So I have to incorporate that also. And related to the artifacts removal techniques like prepossessing that was reduced and compared to producing machine learning techniques, the conversion network was not that much.

And and needed that much pre processing and also that feature extraction task is also reduced. So the conversation network was extracting its own features trying to classify and create performance.

So transitioning to traditional machine learning techniques to confusion on network was easy.

That in Group, the performance and that training process. Of course, the training process may take longer time and resources, computational resources it can utilize more, but the accuracy was given and the performance was also in.

Understood, Professor. You highlighted essential aspects such as the reduction in preprocessing requirements and automated feature extraction with convolutional neural networks, which led to improved classification performance despite higher computational demands. One follow-up: How did you ensure robust generalization of your convolutional neural network across subjects with varying EEG patterns, since EEG signals are significantly heterogeneous?

Uh, yes, that's, that's an important aspect. Uh, so when we observe the, uh, classification performance across different subjects, uh, it varies actually for a model that works perfect for one subject was not working that much, especially for other subjects. So, uh, the approach I was using was the transfer learning technique. So if, uh.

Uh, UH-1 model is trained for one subject, uh, specifically with that data, Uh, if we are applying that model to someone else, uh, we can use that train model and transfer it with a smaller number of data from the other subject. So in that way we can improve the accuracy and I'm sure generalization otherwise, uh, at the universal model for that works for all subjects, uh, is still a challenge. So, uh.

To work more on that, but right now the only way I see is to is transform.

Thank you, Professor. Your approach with transfer learning to address subject variability is both practical and insightful, especially given the challenges of achieving universal generalization in EEG data. Let us proceed to the next skill area. Now, about your ability to teach theory and laboratory courses: Could you describe your methodology for structuring a newly introduced theory course to ensure both conceptual clarity and practical relevance for students?

So for newly introduced theory subjects, you know like.

1st, we have to prepare the study material for that, we have to look for the, uh, price object the student has already learned. So from that concepts is we have to use, we have to adapt that concepts for the new subject. So the first approach I'll do is I'll, I'll make the lecture notes.

Uh, and then some lecture materials like PPPs and diagrams and all, and then we have to arrange it.

In a progressive manner. So the first thing that need to that a student should only learn. We need to teach it first. So that progression should be the.

So, uh, of course for a subject, MPC levels, social levels should cover the topics, uh, which should be learned for that particular course. So the first module should be the foundation. And when we're teaching foundations, we must ensure that the student knows the prior basics so that we have to make sure that student already know some basics. If we, we, we feel that the student lacks that prior knowledge, we have to, uh, give some.

And, uh, some additional materials for covering that as so since it is a new subject, I don't do that much, uh, materials available for the students to, uh, refer directly, uh.

So the questions also, uh, that it shouldn't also, uh, must need to perform well in the exam also. So we should be aware of what kind of questions, what kind, what should be the pattern for that testing interference. So that goal, uh, we should train that students to tackle that boosting vehicles. So, so that.

You should follow that syllabus strictly and in a progressive manner. While the student should be able to learn and uh, we should give assignments also.

Uh, whatever he has learned, uh, for the module, we have to give some practical, uh, applications. So assignments that he, a student should, uh, practice. And these cities, uh, more mathematical oriented subject, we have to give a lot of problems in Class 1 and also less as, as assignments that the students should also and move it problems. So if it is a practical oriented subject, then we have to give some project.

To execute for the student and if it is a theory oriented subject, uh, we have to give some lecture notes and materials that shouldn't students can learn further read and learn. So that should be my efforts.

Thank you, Professor. Your approach to developing a theory course demonstrates a clear understanding of course structuring and the importance of scaffolding material to match students' prior knowledge. Could you elaborate on how you ensure students remain actively engaged during theory classes, especially for subjects that might seem abstract or highly technical?

Yeah, that's, uh, that's it. Very challenging part of teaching.

MMM.

So, uh.

We have to uh.

Actually, we have to, uh, engage them more, uh, with problems. Uh, if that's a few theory subjects, we have to engage the miss practical explanations and examples that they can relate. And we have to in the 1st 1020 minutes, they'll be actively involved in the class.

Take your

And we feel that students are not listening that much. We have to host for some time and we have to.

Do some activities that students regain their energy and concentration, Then we we can go ahead with the next subject. So that shouldn't should be able to relate to the subject. So we have to give more examples and we feel students are not catching up. We have to repeat the topics we already said. Then we can ask questions randomly to send students and encourage them to participate in the class interactive.

Participation is also really encourage them to listen and understand themselves.

Your emphasis on interactive engagement and practical relevance to sustain students' focus in theory classes is very thoughtful, Professor. Encouraging participation and incorporating relatable examples are indeed effective strategies. Let us now discuss your experience in student evaluation and exam duties. Could you describe your approach to designing fair and comprehensive assessments for a course, ensuring alignment with its learning objectives?

Yes, actually we should. We will have some course objectives for every course. There will be some objectives for each question we have been mapped with that course objective.

Uh, and there'll be some, uh, uh.

The overall objective of that.

Program on cities will be there so that program objecties will also emancipated post objectivity. So for when we setting up to the same papers we have to ensure that each question will be peak.

Thank

Corresponding to some of the course objectives, and then we have to consider all the students.

The the question should be uniform like and we have some higher level questions that students should think more and will be something will be difficult to answer and there will be some moderate level testing some basic questions also.

You make the students feel comfortable by writing the exam, so that will be not.

Understood, Professor. You emphasize aligning exam questions with course and program objectives while maintaining a balanced difficulty level to ensure a uniform and fair evaluation process, which is critical for effective assessments. When evaluating exam answers or assignments, how do you ensure consistency and objectivity in grading, especially for more subjective or open-ended responses?

So that, uh, I said, professor, we should not, uh, be biased, uh, to a particular answer that we have in our, in our mind, the students might be, uh, like smarter than the professor itself. So we have to be open minded and at least we, uh, save some question, please, uh, how to, uh.

Uh, we, the concept we have learned, uh, always we have to, uh, be open minded and try to learn from students always in some work questions. So we have to refer and make sure that the student is fight or wrong. If is wrong, we have to correct it in a costly way. And if it is right and how understanding is wrong, we have to appreciate the student and we, we should, as a professor, we should.

Uh, learn that and.

And if it is a question is very often ended, then we have to be often also.

Propagate to the next level of students.

Thank you, Professor. Your emphasis on maintaining an open mind and objectively evaluating student work, even in cases of open-ended responses, underscores your commitment to fairness and academic growth for both students and faculty. We'll now move to your ability to guide student projects and research. Could you share an example of how you have advised or supervised a student project, ensuring both collaboration and the student's independent learning?

Actually when we are guiding, when we guiding projects.

Sorry.

No problem, Professor. Please proceed when you're ready.

OK, so when we are guiding students, we should motivate them to think beyond the out-of-the-box because they should have a problem with them and they should have a solution in mind and they have to experiment on that solution and if it is not working, they have to.

Try and move with some other solution. So I said guide the PT will be to motivate them. Motivate them when they are feeling discouraged.

So and guidance give guidance on on how to.

Work for that. So without without giving specific ideas, we could give them ideas that they can think and they can work on them. Yeah, they said come up their own products and ideas. So that should be the approach in guiding, especially the guide should encourage them and help them to navigate through different. Cars the guy should give them.

I've contacted the they can contact some advice so that's either role of the day.

Thank you, Professor. Your approach emphasizes empowering students to think independently, experiment with solutions, and persist through challenges, all while providing guidance and encouragement—a vital quality in mentoring academic projects and research. Next, let’s examine your communication skills and structured teaching approach. How do you ensure clarity and effectiveness in your verbal or written communication when teaching highly technical concepts to students?

OK. So for verbal communication?

We have to maintain slow, uh, pace of, uh, speed when we log it, we have to, uh, speak clearly if there's one figure. And then in return communication, uh, should be simple. The all kind of students should understand that no matter to which proficiency of the student, you should invest in this so that we should keep it single, uh, very much.

Verbal or, uh, writing also?

Please

That should be the please.

Thank you, Professor. Keeping communication pace steady and simplifying complex concepts for students at varied proficiency levels ensures inclusivity and understanding—qualities essential for effective teaching. Let us move to your academic qualifications. Could you share more about your PhD specialization and how it aligns with the role's required expertise areas?

Thank you. So for my PhD, I was working at the Department of Electrical Engineering my.

What was the diabetics universe? My work was based on the biomedical skin processing lab in that department. So in our lab mainly biomedical equipments like easy measurement and ECG measuring EMC like 5M signal at his foot taking setups further, I was working on UG signals so.

Go on,

I applied signal processing techniques for uh electronic physicians so that it's a file identical 0.

Thank you, Professor. Your PhD specialization in biomedical signal processing and its focus on EEG classification is highly relevant to the expertise areas required, particularly in artificial intelligence and health informatics. Let’s now move to research publications. Could you briefly describe your experience with publishing research in reputable journals and highlight one of your most impactful works?

OK, so, uh, uh, I worked 2 journals, uh, which are sci English science citation index, uh, indexed journeys of, uh, good citation index. So, and I have one quick chapter which is also published by repeated obligation takers and Francis. And one of my journal was published by Springer and the other was Wiley and I had uh.

Publications in IEEE Scopus indexed conferences and I was doing my. When I was doing my Mtech, I published another international conference title track 2IN IEEE conference in the national conference and the two of my.

Uh, do a conference at.

Papers are actually under reviewing, I mean to international conferences. So it's important to publish the paper and input the paper journalists and conferences which is psyched indexed in Group.

Secondary.

Would index vaginas like sign citation index, office inputs or I look for?

Thank you, Professor. Your experience with publishing in high-impact journals like Springer and Wiley, along with IEEE conference presentations and book chapters, demonstrates your commitment to advancing your research within reputable scientific platforms. Regarding industry projects or consultancy, could you describe any collaboration with industry or practical consultancy work related to your research expertise?

As letter to my research, I have been collaborated with in industrial and work at science.

So industry collaboration is not is not the doing my PhD, but I work in one for two years in energy management center, which is actually at that time I had chance to collaborate with various industries in terms of energy conservation, energy auditing that energy related.

Industries.

Thank you, Professor. Your work at the energy management center and collaborations with industries on energy conservation and auditing reflect valuable exposure to applying technical expertise in a practical context. All required skills have now been reviewed. Thank you for your detailed responses and reflections. Could you confirm if we may conclude this interview, or is there anything additional you'd like to add?

Nothing we can convert in that.

Can I confirm that you’d like to end the interview now, Professor?

Yes, we can end the interview.