July 21st, 2023
I arrived in Massachusetts by Uber, sharing the ride with a fellow student who was also moving there for his studies. We pulled up at 10 AM, leaving me with just two hours to spare before our scheduled meetings at noon. I knew I still had a 1-hour walk ahead of me.
As I approached the impressive MIT buildings, I went straight for the Admissions office. And there, sitting before me, was a familiar face.
"Hey, you're the Aliens guy from that lecture!" I exclaimed, my surprise evident.
The blonde guy looked a bit sheepish. "Ah... yes, hi," he replied, visibly embarrassed.
"You had K2-18b on your mind, didn't you?" I inquired.
His face lit up. "Yes!" he exclaimed as if I had just restored his faith in humanity.
I chuckled. "Well, you could have picked a better time to discuss it, though. By the way, what's your name?"
"I'm Rick Rosby," he introduced himself. At this point, I expected he would ask for my name in return, but he remained silent, not giving the impression of being the sharpest tool in the shed.
"I'm Max Sullivan," I offered. "Are you here for the scholarship meeting too?" I assumed that Rick had been invited due to some sporting achievement, given his impressive physique.
"Yes, I'm applying for Software Engineering,"
"Oh! That's interesting," I remarked. "We're applying for the same major, then. I suppose we're going to be classmates." This revelation surprised me; I had initially pegged him as a candidate for the Sports Science department or something similar.
"Let me add you on Messenger, Rick. You cool with that?"
"For sure, for sure," he agreed, fumbling to retrieve his brand-new Samsung phone.
"Is there anyone inside the office right now, Rick?"
"Yeah, there was this one girl who went in a couple of minutes ago," Rick replied, still glancing through his phone - "Is this yours?" - He then held up his phone, showing me the invitation I had just sent.
"Yep, that's me," I confirmed.
"So, Rick, sorry for assuming, but are you an athlete?"
"Everyone seems to think that," Rick chuckled. "Actually, I'm here because I won the Computer Science Olympiad last summer. But keeping your body in good shape helps the brain, you know?"
We got to chatting about computer-related stuff for a hot minute, and let me tell you, Rick had a way of geeking out that left no doubt he was a tech nerd. Now, when it came to the social game, well, Rick might've been a bit short in the street smarts department.
A few minutes later, the office door swung open and out stepped a girl with fiery red hair. She sported a pair of chic glasses and had a radiant smile. Her figure was slender, and she was dressed in sharp, professional attire. A pristine white blouse and a black skirt falling just below the knee.
She briefly glanced at us, nodded, and then gracefully turned on her heels and walked away.
Moments later, a voice from inside the office called out, "Richard Rosby, please!"
Now I was chilling on my own with my thoughts, just waiting for my turn. Gotta say I was feeling pretty confident in my prep. Two weeks back, I started off by grabbing tons of meteorological data and diving deep into the latest statistical science. Especially in that second week, I began to spot some patterns in the historical weather data and used statistical methodologies that were made for the job.
The app wasn't fully done yet, but let me tell you, the results from the mock-up were pretty darn impressive. I could probably score some serious cash just for the theoretical groundwork.
As I sat there, getting my thoughts in order, Rick strolled out of the office room, with a poker face.
"Good luck," he tossed my way before sauntering off.
"Maximilian Sullivan, please!"
I rose from my seat and stepped into the office, greeted by a scene that instantly weighed on my nerves.
Three chairs were neatly arranged on one side of the desk, while a lone chair faced them from the other side. The jury sat in the three chairs, their expressions gravely serious, but one seat remained empty.
"I'm late," - at that moment someone entered the office and I knew who it was, Professor Milik
"You thought I would just leave someone I gave a recommendation to, for Them to eat?" Professor Milik's voice boomed as he entered the room, his words laden with a mix of authority and annoyance. His gaze swept across the jury, daring them to protest his presence.
The jury, though visibly displeased by Professor Milik's uninvited appearance, reluctantly gestured for him to take a seat. As a result, there were now four jury members seated opposite of me.
"I'm Dean Etnis Jones, please seat Max," a man who appeared to be in his fifties announced, motioning towards the chair.
"We welcome you, Max, to MIT," Dean Etnis Jones said with a warm smile that eased some of the tension in the room. He then proceeded to introduce the members of the jury.
"This is Professor Liam Osby of the Software Engineering department," he said, gesturing towards the man seated in the middle. Professor Osby, who appeared to be the youngest among them, nodded in acknowledgment.
"And Professor Emma Peer, responsible for Scholarships and Financial Aid," Dean Jones continued, indicating the older, experienced woman who sat on the jury. Professor Peer offered a polite nod, her expression neutral yet attentive.
"Max, we're here today because you've been offered the opportunity for a Full Scholarship. Your academic achievements have been exceptional, and with the strong recommendation from the Head of the Math Department, we've decided to extend this opportunity to you," Dean Etnis Jones explained. He leaned forward slightly, his gaze fixed on me
"Now, please tell me, when did your interest in mathematics begin?" Dean Jones inquired, his tone earnest and genuinely interested in my response.
"Two months ago," I replied sincerely, and the jury exchanged glances, clearly taken aback by this short timeframe.
'Well, shit... who wouldn't be surprised
"Did you manage to learn statistics this proficiently in just two months?" Professor Milik inquired, his skepticism evident in his tone as he questioned my rapid progress.
I decided to shift the conversation's focus to my project, realizing that otherwise, I might not get the chance to get to it.
"Yes, I was working on developing a predictive model for one of my applications," I explained. "That's what initially sparked my interest in mathematics and artificial intelligence."
"What sort of application are we discussing?" Professor Liam's curiosity was piqued.
"It's a Weather Prediction App,"
Professor Liam raised an eyebrow. "You're aware that there are already numerous tools and extensive research in that field, right?"
"I'm aware of the existing tools, but none of them boast a 98% accuracy up to 9 days," I stated confidently
Professor Liam's interest was piqued even further. "Are you suggesting that your method surpasses anything currently available on the market?"
"Yes it does"
The jury exchanged surprised glances and hushed whispers, clearly taken aback by my bold claim. That's when Professor Milik whispered something to Professor Emma Peer
"Max, could you show us the math behind what you have created? We have no right to use any of your ideas, so provide us with as much as you feel appropriate. We would like to see what you're capable of" - Professor Emma asked
"Of course,"
"The whiteboard is yours" - Professor Milik exclaimed with visible anticipation
I retrieved a stack of papers that I had printed earlier that morning, containing statistical data showcasing the remarkable results of my weather prediction program. With a confident demeanor, I handed the documents to the members of the jury. Then, I rose from my seat and approached the whiteboard.
The jury observed with interest, eager to witness the details of my seemingly groundbreaking work.
I slowly began, "By understanding the underlying temporal dependencies and seasonality patterns in historical weather data, I sought to gain an edge by analyzing the time series. I wanted to discover all the dynamic relationships between multiple variables over time"
Professor Milik nodded as the jury listened in silence.
"I focused on the autoregressive model which predicts a variable's future values based on its own past values, but of course, I needed to extend this to multivariate analysis, so that each variable in the system is regressed on its own past values and the past values of all other variables in the system."
At this point, I started writing equations on the whiteboard
'For each variable Y_i, the WAR(p) model can be expressed as: Y_i,t = c_i + φ_(i,1) * Y_i,t-1 + φ_(i,2) * Y_i,t-2 + ... + φ_(i,p) * Y_i,t-p + ε_i,t'
"So far, what you've described is the standard model used in modern weather predictions," Professor Liam stated, leaning back in his chair. "The real challenge lies in finding an effective method to estimate these coefficients accurately."
The jury members nodded in agreement, acknowledging the complexity of the task. It was clear they were eager to understand how my approach differed from conventional methods.
"That's true, various techniques have been used before, like ordinary least squares or maximum likelihood estimation. The choice of the lag order p can be determined using statistical tests like the Akaike test or the Bayesian test"
"But my approach is slightly different, from the very beginning I established an AI engine calling it PT-GAI, Pre-trained Genetic AI"
"To optimize the way the coefficients are estimated I began to expand the number of variables. I used the Seasonal decomposition of time series adding variables like the trend component, and seasonal component which represent the regular patterns in the data. As well as the residual component."
"On top of it, I used Granger Causality to identify causal relationships among different weather parameters, allowing my models to capture the ripple effects of changes in one variable on others."
Professor Lucas's eyebrows raised slightly, and it was evident that my innovative approach had captured his attention.
"With X representing the causal variable and Y the effect variable, GC provided a quantitative measure of causality"
'GC(X→Y)=F(Yt∣Yt−1,Yt−2,...,Yt−p,Xt−1,Xt−2,...,Xt−q)'
'GC(Xi → Xj) = ∑{k=1}^{∞} (θ1k ⋅ X1-k)'
"p and q are lag parameters, Yt represents the effect variable at time t and Xt represents the causal variable at time t"
At this point, the jury was truly really impressed and even Professor Milik had a smile on his face.
"With a great inflow of data, I ended up with over a thousand different variables, and all I needed to do now was find the coefficients that gave the most accurate results. That's where PT-GAI came into play. I needed to spend some money on renting out a more powerful server because my PC was heating up."
I paused for a moment, and then continued, "I established an initial population P0 of diverse models, each represented as a chromosome and each having different combinations of the coefficients"
'Chromosome_n = Θ = {θ1, θ2, θ3, ... , θn}'
'P0={Chromosome1,Chromosome2,...,Chromosome_n}'
The jury exchanged impressed glances, recognizing the sheer magnitude of variables and computations involved in my work.
"Then," I went on, "I created a fitness function which quantified the performance of each chromosome by measuring its predictive accuracy concerning historical weather data."
"Finally," I concluded, "I implemented selection, crossover, and mutation processes which governed as to which forecasting models are favored or changed over iterations, but I would prefer not to disclose those."
The room fell momentarily silent as the jury absorbed the magnitude of my achievement. It was clear that they were witnessing a groundbreaking approach to weather prediction.
Professor Milik leaned forward and spoke, his voice filled with admiration, "Max, what you've presented here is nothing short of remarkable. This is certainly a cutting-edge fusion of statistical science and AI. We look forward to using this application in the future."
The other members of the jury nodded in agreement, their amazement evident in their expressions.
Professor Milik, after a moment of contemplation, leaned in once again, his eyes twinkling with enthusiasm. "Max, your talent and dedication are truly exceptional, especially for someone your age. I have a proposition for you. How would you like to represent our institution in the upcoming Math Olympiad? Your innovative thinking and problem-solving abilities would make you a valuable addition to the team."
I was taken aback by the offer and the other jury inside the room were even more surprised, but a moment later their expression changed as if it was only natural.
To be recognized and invited to join the Math Olympiad team at MIT as a freshman was a great honor.
"I agree" - I nodded my head.
Professor Emma Peer, with a warm and reassuring smile, leaned forward and addressed me directly, "Max, based on what we've witnessed here today and your exceptional accomplishments, I think there is no point in further discussions. I am delighted to inform you that you are granted a Full Scholarship with Additional Aid for living expenses during your four years at MIT. The total amount allocated to you is $380,000"