Machine learning halide Perovskites - Finding the Goldilocks Zone

Before jumping on to "Machine Learning halide Perovskites", let's understand first the two terms "Machine Learning (ML)" & "halide Perovskites".

Perovskites are an organic/inorganic hybrid material used to make solar cells (PSC- Perovskite Solar Cells) and other optoelectronic devices, which turn light into electricity. They are extremely efficient and inexpensive to produce and therefore, to the researchers, perovskites are the future for renewable energy to make solar power accessible and affordable. However, they are highly unstable materials and can quickly degrade under a range of conditions including light, temperature, humidity, oxygen and electrical bias.

Machine Learning (ML) can be defined as the subfield of artificial intelligence that involved the development of self-learning algorithms to gain knowledge from that data in order to make predictions. Instead of requiring humans to manually derive rules and build models from analyzing large amounts of data, machine learning offers a more efficient alternative for capturing the knowledge in data to gradually improve the performance of predictive models, and make data-driven decision. In layman's terms, when our brain recognizes any object, the answer comes from the "knowledge" that we acquire. For instance, when you see an image of a pizza and your brain recognizes it, then the answer is coming from the "knowledge" that you acquire. You may have smelled pizza or have tasted it. Similarly, when a naive machine (it may be a computer, an application or website etc) is trained to make decisions then it is called machine learning.

Applying ML to halide perovskites is just like finding the "Goldilocks Zone" of appropriate conditions because there are so many factors which can degrade perovskites independently or in combination. Researchers implement ML on halide perovskites to investigate stability both at material and device levels, to check the formability of material and to understand the hidden laws governing the properties of this class of material. To learn more on this topic check out this article: https://doi.org/10.1016/j.jechem.2021.07.020

Ph.D. (People Have Doubts)

As I have completed more than 2 years of my Ph.D., the following are some of the things which I have always tried to follow and some of the things I wish I had known earlier.

• Avoid Comparison: Don't try to compare yourself with your peers. The context and the journey matter. Striving to be the best creates an illusion of an endpoint and therefore, shift your focus to be at least 1% better each and every passing day than your past self.

• Be a critical thinker: My supervisor has always wanted us to be critical thinkers. Don't be a doer rather think first. Half of the work comes from identifying a compelling problem. Spend more time reading and thinking than doing. There is no point in rushing down a project and then realizing there is no point in pursuing it. This is especially difficult in academia's publish or perish culture.

• Learn to handle failures: In this long journey, failure is as certain as day and night. Progress and productivity come in waves. When your experiments and research is going strong, capitalize on it. When you have experienced failure, analyze it and learn from it. When you have hit a wall, don't bang your head against it. Take a step back and re-access the situation.

• Being Selfish: Prioritizing yourself and your work is NOT selfish. Learn to say no to extra responsibilities, side projects, even mentorship opportunities. Overcommitting to the promises which you can't keep eventually is SELFISH. You cannot pour out of an empty cup.

• Developing Skillset and Begin with the end in mind: PhD is a time to become specialised and a time period to DEVELOP skills. More than 60% of candidates in PhD want to pursue career in academia, but not all of them end up pursuing it. Diversify your skillset. Set broad goals and leave as many doors open as possible.

• People Matter: In success, there is always "US" and not "ME". Great research comes from being part of a great team. Find advisors, mentors and peers who genuinely supports your growth. Surrond yourself with people who will be there for you in your rough times of experiment failures, paper rejections and downtime.

• Health First: PhD is a degree that needs to be earned and hence it requires hard work and dedication. However, you are above your research. Find hobbies outside of research through which you can go to a state of relaxation and flow. Taking breaks means taking your mind 100% off work as much as possible. Practice mindfulness.

"Consistency is always better and greater than Perfection."