Discussion on the Development of a Graphical User Interface for Time Series Toolbox Using Deep Learning

[sajontahsen]

Hello! I'm excited about the possibility of contributing to this project and have a few queries:

1. Contribution Opportunities: As a newcomer, I'm eager to understand how I can begin contributing. Are there any tasks or issues suitable for someone new to the project?
2. Datasets and Tools: For experimenting with time series data and deep learning models, are there any recommended datasets or tools that I should start exploring?
3. GUI Frameworks and Technologies: I noticed the selection of Gradio. Could you share the reasons behind this choice, particularly any specific advantages or limitations? Additionally, given the versatility and integration capabilities of Flask or Django, is there room to consider them as alternatives for this project?

I'm looking forward to your guidance to better frame my approach to the project prototype.

Best regards,
Tahsen Islam Sajon
Resume | Email | LinkedIn

[aabdelre]

Hello,

I am super excited to contribute to this project. I understand it is a new project, but I was wondering if you are expecting potential GSoC contributors to complete some tasks to be eligible to apply. I was also thinking, for this project, to try out an LSTM network architecture. LSTMs can better capture seasonal trends and dependencies over longer periods compared to RNNs. These things are always task-specific and it's impossible to know which one would perform better, so it is definitely something worth looking at. I also wanted to know about expectations from the contributor by the end of the project. Will the project be deployed somewhere or will it only be used locally for research purposes? Please let me know as it would greatly help me with my proposal. Thanks for your time and consideration!

Sincerely,
Ahmed

[A-SOLO]

Hello there,
I am Anupam Rai, student at IITBHU. I am interested in this (Development of a Graphical User Interface for Time Series Toolbox Using Deep Learning) project too.
I have extensively worked on image segmentation techniques. And I have experience with Python, TensorFlow, PyTorch, and PyQt5.
I also have experience in Time Series data through my recent hackathon and projects.
I look forward to collaborating and contributing to this project.

Thanking you,
Anupam Rai
anupamrai481@gmail.com

[Debangan-MishraIIIT]

Greetings,
I am Debangan Mishra, a 2nd year undergraduate student at IIIT Hyderabad. I am eager to work in this project (Development of a Graphical User Interface for Time Series Toolbox Using Deep Learning) as I have prior experience in building machine learning tools for analysing time series. One of my projects deals with financial time series prediction using CNN-LSTM network (a paper implementation). I have a strong interest in research and learning new concepts, particularly in the field of machine learning (as demonstrated by my project on multimodal hateful meme detection, among others)

I am comfortable with PyTorch, numpy, matplotlib, Python, C/C++, Javascript, MERN stack, etc. and am eager to work on this topic as it would help me greatly develop my skills in these fields.

My resume is attached below:
https://drive.google.com/file/d/1wItKJaVLfTRi6ivxuTxICcMQWiwuc-MI/view?usp=drive_link

[jaygodara45]

Respected sir,

My name is Jay Godara, a third-year undergraduate student at the Indian Institute of Technology Kharagpur.

I found this project relevant to the research I am working on at IIT Kharagpur under Prof. Mahesh Mohan MR. We have been exploring different time-series models. We are going deeper into TimesNet, which has proved to be a better model than Transformers/CNNs for time-series data.

My proposal:
The TimesNet model can be integrated for better accuracy.

My experience with time-series data:

1. Currently working on a research project regarding time-series data | Prof. Mahesh Mohan MR
2. Smart Traffic Management System | Prof. Sudip Misra
   • Developed an LSTM model for traffic volume prediction, achieving over 95% accuracy
   • Utilized computer vision knowledge for real-time vehicle detection

Other relevant skills/experience:
• Proficiency in PyTorch, TensorFlow, OpenCV, and other ML/DL libraries
• Proficiency in Python, C++, C
• Worked as a design lead at a college startup: ROGAN, developed the complete UI/UX of the application.

Resume link: https://drive.google.com/file/d/1Uf3DBcEeT0YL19imDi1oinOEALLAlPTp/view?usp=sharing

Being from the Biotechnology department and enthusiastic about Artificial Intelligence, I consider this project a proper fit for me. I would be happy to contribute.

Thanks,
JAY GODARA
jaygodara@kgpian.iitkgp.ac.in
jaygodara10062006@gmail.com"

[Precious-EEE]

Hi
I am Adekunle Precious Oreoluwa,I am 500 level student of Obafemi Awolowo University,I am interested in this project "Development of a Graphical User Interface for Time Series Toolbox Using Deep Learning".
I have interest in cutting-edge technology and data analysis which aligns perfectly with this project. I am proficient in Seaborn, matplotlib, PyTorch, and different deep learning libraries.

Thank you
Adekunle Precious Oreoluwa (preciousadekunle33@gmail.com)

[jasonxiexy]

Hello,
My name is Jason Xie, and I am a Meng ECE student currently at University of Toronto, and undergraduate studies in computer science at Western Univeristy. I am excited about the opportunity to contribute to the development of a Graphical User Interface (GUI) for the Time Series Toolbox using Deep Learning under the mentorship of Ozgur Kara and Mahmoud Zeydabadinezhad.

Having focused my studies on algorithms, machine learning, and related software engineering skills, I am particularly interested in contributing to projects that involve the implementation of machine learning algorithms, specifically deep learning on medical time series. I believe my background and skills align well with several aspects of this topic, including:

1. Deep Learning Expertise: I have a strong background in deep learning techniques, including Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs), which are essential for analyzing time series data. I have worked on projects involving neural network architectures for various applications, including Generative Adversarial Model on image generation.

2. GUI Development: I am proficient in developing user-friendly GUIs using tools like Qt5, or web-based frameworks like Vue. I understand the importance of creating intuitive interfaces that simplify complex data analysis tasks, making them accessible to users with varying levels of expertise.

3. Data Handling and Preprocessing: I have hands-on experience with data preprocessing techniques, including cleaning, normalization, and handling missing data. I am familiar with tools such as Pandas, NumPy, and Scikit-learn for efficient data manipulation and preparation.

I am committed to delivering a functional and user-centric GUI for the Time Series Toolbox, integrating deep learning models, robust data handling capabilities, and interactive visualization tools. I look forward to collaborating with the team to achieve our goals and make impactful contributions to the project.

Thank you for considering my candidacy, and I am eager to contribute my skills and expertise to this exciting endeavor.
Jason Xie
xjason0312@gmail.com
[]
(https://www.linkedin.com/in/jasonxie0312/)

[atiabjobayer]

Hello, I am Atiab Jobayer. I am a bachelor student from NIT Tiruchirappalli, India. I am really eager to contribute to this project. I have decent experience in both deep learning and UI development.

I have previously worked at Samsung and Goldman Sachs. At Samsung, my work was on developing deep learning model for normal map generation - which makes me a very eligible candidate for this time series UI development project. I have a thorough understanding of the models.

In terms of UI - I have 6 years of experience in web development, for that I have developed a lot of UI from scratch using plain HTML CSS and UI frameworks like React. I also have little experience with QT5.

Thank you for considering me and I am really looking forward to working with you.

LinkedIn: https://www.linkedin.com/in/atiab-jobayer/
Resume: Resume.pdf
Email: purnoak2013@gmail.com

[Iamkrmayank]

Hi there,
From Graphic Era University in Dehradun, I'm Kumar Mayank. Development of a Graphical User Interface for Time Series Toolbox Using Deep Learning is a project that interests me as well.
I've done a lot of research on image segmentation methods. Additionally, I've worked with PyTorch, TensorFlow, and Python.
Time series data is another area in which I have expertise thanks to study projects.
I am excited to work with you on this project and add to it.

Thanking you,
Kumar Mayank
krmayank2002@gmail.com

[Divyam6969]

Hello there :) 👋

I am Divyam Malik, currently in my third year of computer engineering at Thapar Institute of Engineering and Technology, Patiala. My journey in software development has been focused on Python programming and deep learning, with a specialization in PyTorch. My drive stems from the belief that technology should be harnessed to tackle intricate challenges and make advanced machine learning accessible to all.

Possessing a strong grasp of software development principles, including version control and debugging, I am equipped to deliver robust and scalable solutions. My expertise extends to image segmentation techniques, particularly in medical imaging, where I have explored diverse datasets like MRI scans. This project resonates with my interests, aligning with my ongoing MITACS Globalink Research Internship, which underscores my commitment to research-driven solutions.

Throughout my academic journey, I have actively contributed to projects that bridge the gap between technology and healthcare. Notable examples include my work on pneumonia detection using FastAI and projects focused on UI/UX design. These experiences have honed my skills and fueled my passion for innovation. You can explore my GitHub portfolio to delve deeper into my projects and contributions.

Portfolio: https://divyam6969.github.io
Linkedin: https://www.linkedin.com/in/divyam-malik-8402a821a/
Github: https://www.github.com/divyam6969
Email: divyammalik2003@gmail.com

[jaygodara45]

Hello! I'm excited about the possibility of contributing to this project and have a few queries:

1. Contribution Opportunities: As a newcomer, I'm eager to understand how I can begin contributing. Are there any tasks or issues suitable for someone new to the project?
2. Datasets and Tools: For experimenting with time series data and deep learning models, are there any recommended datasets or tools that I should start exploring?
3. GUI Frameworks and Technologies: I noticed the selection of Gradio. Could you share the reasons behind this choice, particularly any specific advantages or limitations? Additionally, given the versatility and integration capabilities of Flask or Django, is there room to consider them as alternatives for this project?

I'm looking forward to your guidance to better frame my approach to the project prototype.

Best regards, Tahsen Islam Sajon Resume | Email | LinkedIn

Some standard and widely used time series datasets are ETT (Electricity transformer temperature), SWaT, SMAP, MSL, EEG, etc. Select one of these and try to train on different deep learning models like Autoformer, Crossformer, TimesNet. Feel free to ask me if you have any doubt. Most lf these datasets are publicly available. For SWaT dataset, you will have to request access and you are not supposed to share SWaT dataset publicly.

[jaygodara45]

For GUI, i think Gradio along with PyQt5 or Tkinter is sufficient. For backend, we can go with Django as it is a python framework... would be easier to integrate.

[jaygodara45]

For anyone reading this comment, would love to know your interpretation of the idea.