Individual retrospective Milestone 4: Communicating Results

[Mohamedmxz]

Please share your individual learning retrospectives for this Milestone

🧠 What you learned

✅ What went well

⚠️ What could be improved

[FalaqMajeed]

🧠 What You Learned
Explored a variety of website AI agents that greatly assisted us in building our websites, opening up an entirely new world in web development.
✅ What Went Well
Reached a smooth and mutual agreement on our project direction, clearly identifying our audience, and forming a solid vision for how we will craft our research.
⚠️ What Could Be Improved
The group should maintain higher engagement and provide updates at every step, rather than relying solely on meetings. I hope that in the final milestone, all members will actively collaborate on the final presentation

[Adamx090]

🧠 What you learned

• How communication medium choice (in this case, a website) impacts accessibility, reach, and the way data is interpreted.
• How to tailor messaging for a highly specialized audience by focusing on evidence-based content, clear data visualization, and professional tone.

✅ What went well

• Smooth team coordination on dividing research, writing, and technical tasks.
• The final site successfully conveys key messages and provides a professional, credible impression suitable for organizations like IHME and WHO.

⚠️ What could be improved

• Seeking feedback earlier from individuals in our target audience to better validate content and design decisions.
• Improving my technical skills in building web interactivity so I can contribute more directly to future data presentation features.

[linahKhayri]

1. What I Learned
   I learned how to merge scientific content, storytelling, and web design into a cohesive platform and I gained new skills in the web development field (html/css) with the help of LLMs and found it really interesting, while also deepening my understanding of air pollution data and health impacts.

2. What Went Well
   We created a website that blends data-driven insights with human-centered storytelling. Features like the PM2.5 size comparison and showing why it matters and why it was our focus brought clarity to complex concepts.

3. What Could Be Improved
   Integrating the prediction tool into the site’s interface would make it much more convenient for users.

[SiSaR-Pal]

🧠What I learned
How to interpret and clearly explain differences between in-sample and out-of-sample R² results.
The importance of aligning features and avoiding re-fitting pre-trained models to preserve model integrity.
Gained experience in integrating cross-technology workflows, specifically the challenge of finding a backend solution with Python support for building a website front end and back end, highlighting practical deployment constraints beyond data science.

✅ What went well
Successfully used four pre-trained Random Forest models consistently without retraining, increasing reproducibility.
Facilitated clearer group discussions by bridging technical results with accessible explanations.
Managed to adapt workflows towards practical application needs, even when encountering challenges like backend support for Python in
web development environments.

⚠️ What could be improved
Incorporate per-category cross-validation or hold-out test splits to obtain more robust subgroup performance metrics.
Explore or experiment with additional backend technologies or frameworks that support Python better, to streamline deployment of data science models into web applications.

[salihakalender]

Saliha Kalender

🧠 What you learned
Tried to learn the essentials of website development.
Understood the importance of reaching the proper target audience for the project’s marketing.
Realized how valuable it is to enrich a website with relevant and engaging content.

✅ What went well
Our team created a functional website where technical analysis can be presented in a way that connects with people.
The team came to a shared understanding of the project goals and audience.

⚠️ What could be improved
We could have decided on the target audience earlier and with more clarity.
I need to further improve my web development skills through consistent practice.

[Mohamedmxz]

🧠 What I Learned
I learned the importance of choosing the right audience for any project and how critical it is to build good communication with them so they truly care about our work and its impact. This isn’t just about sharing facts, but about understanding the audience you’re targeting and being confident in your work. I also learned how to use Streamlit, a Python library for creating websites.

✅ What Went Well
We successfully developed a website for our communication strategy and explored different tools to achieve the best results.

⚠️ What Could Be Improved
We are considering creating a tool to predict DALYs using our machine learning model. We are also thinking about changing the random forest algorithm to improve the model’s performance.