From Data to Understanding: My Journey Through BINDS Chapter 2

Bridging Nature with Data Science Workshop

Azim Premji University, Bhopal | January 29-31, 2026

Three days. That is all it took for something to shift in how I think about data, about learning, and about what makes us human in an age of artificial intelligence.

I just completed BINDS Chapter 2 - Bridging Nature with Data Science - at Azim Premji University in Bhopal. And I want to tell you what happened, because it has changed how I see learning, problem-solving, and the relationship between humans and technology.

What is BINDS and Why It Matters

• The goal of BINDS is simple but powerful: to equip everyone with modern data science tools and technologies. Not just for the sake of learning tools, but to understand that data science is essential across all developmental sectors, modern industries, and research fields.

• In this data-driven revolution, students, teachers, researchers, and professionals - especially those with limited technical expertise - often face both theoretical and applied challenges. BINDS aims to transform us from passive observers into data-driven advocates. The workshop intends to build technical competence while fostering a sense of belonging in this rapidly changing data landscape.

• But more than that, it gives us a highly marketable skill set - the ability to create evidence-based narratives, to understand the world through data, to ask the right questions.

The Instructors Who Made This Happen

First, let me acknowledge the incredible team behind this workshop:

Dr. Abhishek Mukherjee

Associate Professor at the Indian Statistical Institute, Kolkata. He holds a PhD from the University of Florida and has postdoctoral experience from Texas A&M University. His research interests include ecological niche modeling, invasive weed ecology, biological control, and nano-biotechnology applications for sustainable agriculture. He brought deep expertise in ecological data science to the workshop.

Dr. Amiya Ranjan Bhowmick

Faculty member in the Department of Mathematics at the Institute of Chemical Technology, Mumbai. His academic interests span interdisciplinary data science, statistical modeling, and applied mathematics. He actively integrates quantitative methods into ecological and environmental research and is deeply committed to training students in data-driven research and computational thinking.

Ruqaiya Sheikh

Project Research Assistant at IIT Bombay and Visiting Faculty in Data Science and AI programs at the Institute of Science, Mumbai, and St. Xavier's College, Mumbai. She holds a master's in mathematics from the Institute of Chemical Technology, Mumbai. Her research interests include machine learning, statistical modeling, and applications of data science across various domains.

Dipali Mestry

PhD student (DST-Inspire Senior Research Fellow) in the Department of Mathematics at the Institute of Chemical Technology, Mumbai. Her research focuses on Bayesian and computational modeling for ecological and environmental applications. She has expertise in Monte Carlo methods, MCMC, Approximate Bayesian Computation, and dynamic documentation using Quarto - which became one of the most revelatory sessions for me.

These four brought different perspectives, deep expertise, and genuine passion for teaching. They did not just lecture - they engaged with us, answered our questions, and made us feel like learning together was the most important thing happening at that moment.

The Problem I Came In With

• Before the workshop, I had this gap in my understanding. I knew data existed everywhere. In my economics classes, in environmental science, in social surveys. Data is all around us. But there was something missing. I could see the numbers, but I could not really see the story they were telling.

• I could open a spreadsheet and see columns and rows. But I did not know how to ask it real questions. I did not know how to make it speak. I did not know how to understand what it was trying to say about the world.

• And I thought the problem was mathematics. I thought I needed to be a math genius to understand data science. I thought the barrier was advanced calculus, complex statistics, theoretical knowledge that I did not have. I thought the problem was in my brain.

• But I was wrong. And that realization started on Day 1.

  
  

Day 1: Data Preparation, Regression, Classification and Clustering

• The workshop began with a welcome that set the tone for everything that followed. We were told: data science is not about being a genius. It is about being curious. It is about asking the right questions. It is about caring enough to understand.

• We jumped into R - importing ecological datasets from CSV and Excel, learning descriptive statistics, creating data visualizations. Then we learned the hardest part: handling missing values, duplicates, and outliers.

• Here is what surprised me: cleaning data is not mechanical. Every decision you make - whether to remove an outlier or keep it, how to fill a missing value, what counts as a duplicate - is a decision about what story you want the data to tell. This is where critical thinking happens. This is not the boring part. This is the most important part.

• Then we learned regression methods. Not just the formula, but when to use them, why to use them, what assumptions you are making. We learned about feature selection, regularization, and how to interpret results in an ecological context. The case study was about environmental predictors of biodiversity - a real problem that real people care about.

• I realized: A computer could run a regression. But only a human asks: Does this model make sense for the real world? Are my assumptions valid? What could I be missing?

• In the afternoon, we learned classification - logistic regression, decision trees, random forests. We learned how to handle imbalanced datasets, which is important when you are dealing with rare species. The case study was predicting species presence or absence.

• Then clustering: K-means, hierarchical clustering, and ecological dissimilarity measures. How do you group habitats or species assemblages? What makes things similar? These are not just mathematical questions. They are ecological questions. They are questions about the real world.

• What I realized: The tools matter less than the thinking. What matters is understanding what question you are trying to answer, then choosing the right tool for that question.

Day 2: Resampling, Nonlinear Models, Prediction and Reporting

Day 2 started with resampling methods - bootstrapping confidence intervals, cross-validation for model validation. These are not just techniques. They are ways of asking: How confident am I in what my data is telling me? How do I know my model is not just getting lucky?

• Then nonlinear regression. The world is not always linear. Ecological relationships are messy and complex. You cannot always draw a straight line and call it understanding.

• We explored predictive frameworks for time series and spatial ecological data. A case study on predicting species distribution under climate change. This is not academic. This is real. Species are moving. Habitats are changing. We need to understand what is coming so we can prepare.

• And this is where I felt something shift. I was not just learning a technique. I was learning how to think about a real problem. How to collect data about it. How to model it. How to predict what might happen. How to communicate that prediction to people who make decisions.

• Then came the moment that changed everything for me: Quarto.

• Quarto is about dynamic documentation. You write your R code, your explanations, your plots - all in one document. The document is reproducible. Anyone can read it and run your code and see exactly what you did and how you got your answer.

• And suddenly I understood something crucial: analysis without communication is useless. You can build the most beautiful, accurate model in the world. But if you cannot tell someone else about it, if you cannot show your work and explain your reasoning, what good is it?

• Data science is not a solo activity. It is communication. It is persuasion. It is helping someone understand the world better. And that requires something machines do not have: the ability to think about your audience, to simplify complexity, to tell a story that matters.

• This is where I felt the difference most clearly. A machine can run analysis. But a human tells the story. A human decides what matters. A human connects the dots in a way that makes sense to other humans.

• We had practice sessions with combined workflows - data cleaning to modeling to reporting. More R coding exercises. And discussion of further applications. By the end of Day 2, I had touched everything. I had done data cleaning, built models, created reports.

• We also had dinner together. This matters. Learning is not just about the content. It is about being part of a community. It is about talking with other people who are curious, who are struggling with the same things you are struggling with, who get excited about the same ideas.

Day 3: Handling Spatial Data

On the final day, we learned about spatial data - spatial data types, data collection, data cleaning. We learned about QGIS - a geographic information system.

• And I realized something: all of this - the regression, the clustering, the predictions - happens in a context. Data does not exist in a vacuum. Your species distribution exists in specific forests. Your economic data belongs to specific countries. Your climate change predictions matter because they affect specific places where people live.

• We learned QGIS - loading shapefiles, CSV files, web-based maps. Performing basic editing and attribute table operations. Geoprocessing tasks like buffering, clipping, and intersection. Handling data in QGIS.

• This is not just mapping. This is understanding space. This is asking: Where are things? How are they distributed? What is near what? These questions matter for ecology, for conservation, for understanding how the world is organized.

• We learned about making professional-quality maps. Adding legends, scale bars, labels. And here I learned again: making a good map is not about aesthetics. It is about clarity. It is about taking complex spatial relationships and making them visible to someone who has not seen them before. It is translation. It is communication.

• We had hands-on case studies in map-making and interpretation. We made maps. We learned to read them. We learned to tell stories from them.

What I Really Learned

If you ask me what I learned in BINDS, here is the technical list: R programming, regression modeling, classification trees, clustering algorithms, cross-validation, nonlinear regression, Quarto, QGIS, spatial data analysis, cartography.

• But that is not really what I learned. That is what I was taught. What I actually learned was something deeper.

• I learned that you do not need to know tons of mathematics to understand data. You need to be curious. You need to ask questions. You need to care about getting it right. This is what creates a difference between us and AI.

• I learned that data is everywhere in my life. In my economics courses, in the environment around me, in the choices people make. And now I know how to read it. I know how to ask it questions. I know how to understand what it is trying to say.

• I learned that expressing myself through code is a form of thinking. When I write R code to clean a dataset or build a model, I am not just telling the computer what to do. I am thinking through a problem. I am making choices. I am having a conversation with the data.

• I learned that the workshop helps me connect with my courses and life learning. How to understand economics data or any kind of dataset. How to understand the world through a way in which I can express myself by coding. This really makes me something new and full with curiosity.

• And most importantly, I learned the difference between humans and artificial intelligence. An AI can process data faster than me. It can run a thousand models while I am still thinking about the first one. But it does not know why. It cannot ask the right questions. It cannot care about getting the answer right.

What Makes Us Different

• This is the insight that keeps coming back to me. We know data. But how do we analyze and understand it by connecting it with real problems? This is what creates a difference between humans and AI.

• We human beings can think. We can act. We can understand WHY we are doing something, not just WHAT. In this time, we also learn new things and apply them practically with critical thinking. This thing creates the difference.

• AI can compute. AI can optimize. AI can follow patterns at scale. But only humans can think critically. Only humans ask Why? Only humans connect a dataset about biodiversity loss to the forest I walked through last week, or to the climate crisis my generation is inheriting.

• A machine can tell you that removing outliers increases model accuracy by 3.2 percent. But only a human asks: But what is that outlier? Is it an error, or is it a signal? Does removing it make my model more accurate or just less inconvenient?

This is what BINDS really taught me. It was not just about tools. It was about thinking. It was about asking good questions. It was about understanding that data science is a human discipline.

Connecting to Everything Else

• Something unexpected happened during this workshop: everything started connecting. The economics courses I am taking suddenly made more sense. I could imagine analyzing actual economic datasets, understanding what the numbers were saying, predicting what might happen.

• The environmental science I have studied took on new meaning. I could see how to measure biodiversity, how to track habitat loss, how to predict where species might survive climate change.

• My life, my interests, my studies - they all suddenly had a tool. Not just a tool, but a language. Data science became a way to think about and understand the world. That is the real value of this workshop. It did not just teach me R and QGIS. It gave me a way to connect what I am learning in the classroom with how the actual world works.

The Thing About Curiosity

• One of the instructors mentioned something that stayed with me: being curious beats knowing everything. And I think that is the core of what BINDS was really about.

• I do not know all of R. I do not understand all of statistics. I cannot build perfect models. But I am curious. I want to understand how things work. I want to ask questions and find answers.

• And that curiosity is what makes the difference. Because in three days, I learned more than I could have learned in three months of watching videos or reading books. Not because the instructors were magical. But because I was actually engaged. I was thinking. I was asking Why? I was connecting things.

• And that attitude of curiosity, of critical thinking, of asking good questions? That is what I am taking away. That is what will serve me better than any specific technique.

Looking Forward

• I am different now than I was three days ago. Not because I am suddenly an expert in data science. I am not. I am still learning, still struggling with code, still making mistakes. But I am different because now I know that I can learn this.

• I know that the barrier is not my brain or my ability to understand. The barrier is just laziness or lack of curiosity. And those are things I can control.

• I am different because I see data everywhere now, and I see tools to understand it. I see problems, and I see ways to approach them.

• I am different because I understand now that being human, being intelligent, is not about computing faster or knowing more. It is about thinking better. It is about asking the right questions. It is about caring about the answer.

• Thank you to Dr. Abhishek Mukherjee, Dr. Amiya Ranjan Bhowmick, Ruqaiya Sheikh, and Dipali Mestry for this incredible learning experience.