How do you execute a computer science research project?

When I was chemistry grad student, we were using the standard scientific-method to write our reports. Upon entering industry, I started learning about Problem Solving Process (PSP), Kaizen, TPM & other analysis tools. My vision to encounter problems changed, as I worked with engineers, sales, tech-support and financial experts. To ID a problem, consider using the 5-Why's and a tree-diagram to root-cause as many possible variables. Then, diverge to ID the true issue, formulation vs. instrumentation (etc.). The conclusion does not end until the problem has a long-term solution. Please consider using other statistical tools to guide you along the process.

Project Title: Predictive Modeling for Healthcare Outcomes Part 1: Project Introduction "We collaborated with healthcare companies to develop predictive models using data analytics and data science." In this project, we partnered with healthcare organizations to leverage data analytics and data science techniques to enhance patient care and outcomes.

In Earth Sciences we use computer models to simulate highly complex systems, usually too complex to grasp. One of the challenges is to simplify the system in a way that enables it to be represented by a program code but stays realistic and provide information on the research question

Embarking on a computer science research voyage, my fellow explorers, begins with the art of finding a problem. It is the compass that points us towards uncharted territories of knowledge. In the intricate web of technology, problems abound, waiting for curious minds to untangle. Seek challenges that resonate with your passions, for therein lies the spark of relentless inquiry. Explore uncharted territories and decipher the questions that have yet to find their answers. Yet, do not rush this journey. Embrace the patience required to unearth a problem worthy of your dedication. In the stillness of contemplation, you shall discover a puzzle that beckons you, one that aligns with your purpose as a seeker of computer science truths.

Identifying a compelling research problem is indeed the foundational step of any project. In my experience, immersing myself in existing literature and actively participating in seminars has been invaluable. For instance, during my master's program at the University of Rome "Tor Vergata," I became intrigued by the potential of Electrical Impedance Tomography (EIT) for wearable health monitoring. Conversations with experts and peers helped refine this interest into a feasible research question. Additionally, allowing my curiosity to guide me resulted in creating a prototype wristband for real-time LIS translation, aiming to improve communication accessibility.

Always start with PEEL (point, example, elaboration, link) framework when writing literature review section. Point: Find gaps, limitations, weaknesses and underlying assumptions from the existing research works Example: group research works with common attributes be it similar optimisation method, solving the same problem etc Elaboration: elaborate and explain why the example is good or bad Link: Link the gaps to your work and how your work can mitigate these challenges.

After clearly describing the problem scenario, the next step is to review the literature that is published in that domain. This can provide us with a clear route to go by exploring the methodologies, algorithms, and techniques that have been used to address these types of problems. And don't limit literature review to published work only; engage in online forums, blogs, and community discussion as well. Consider organizing your literature review chronologically or thematically. A chronological approach helps trace the evolution of ideas over time, while a thematic approach allows you to categorize research based on common themes or concepts.

I agree that once a research problem is identified, conducting a comprehensive literature review is crucial. In my experience as a mechatronics engineer passionate about robotics and machine learning, immersing oneself in existing literature provides valuable insights into the background and current advancements related to the problem. This process involves critically evaluating previous work to identify strengths, weaknesses, and gaps. By synthesizing these, we can position research effectively, justify methodology and ensure that we are contributing original insights rather than duplicating existing efforts. A thorough literature review not only informs our approach but also enhances the overall quality and impact of the research project.

Make sure to log your results properly in the meantime, regardless good or bad results they are all important. In fact, bad results may show more insights to your approach and therefore you can tune and make it better.

One important part of research is writing the report. Ensuring the report has all the necessary details for understanding, verifying and reproducing the result by a third person is vital. In the domain of computer science, having appendix with details to the tools and softwares used, explanation of theorems and results used etc add value to the research.

Make sure that your work adheres to the format. Ensure that your work flows well, and shows a coherent thinking process. Make good use of spaces and paragraphs to draw readers attention. It is better to use simple language rather than pompous ones which could make reading convoluted. Try to keep sentences short. Usually, If it is more than 20 words long you could probably break them up into separate sentences. It makes reading whole lot easier, especially for long papers like journals articles.

Better be clear what is purely a result of abstraction and what is definitely events of nature in simulating new concepts. In this way, you know what is scientific and what is speculative

Build a proof of concept based on existing knowledge. Test it in real life conditions. Distill the findings and draw inference. Reiterate and monitor system behaviour throughout.