Aaditya Prakash (Adi)
Principal Machine Learning Scientist in Drug Discovery at Genentech - building foundation models and single-cell embeddings for high-content imaging and optical pooled screens. Previously Spring Science (formerly Spring Discovery), PathAI, Microsoft Research, Qualcomm Research and Philips Research. Ph.D. in Computer Science from Brandeis University.
Google Scholar LinkedIn GitHub About & publications Notes adi.prakash.ml [at] gmail [dot] com
Writing¶
Pixel Deflection - a simple transformation based defense¶
February 1, 2018
This classification function works for multiple images in a batch. However, visualization is limited to the first image in the batch. See our GitHub code for better batch processing functions.
Visual Question Answering Demo in Python Notebook¶
April 4, 2016
@media print { div.cell { display: block; page-break-inside: avoid; } div.outputwrapper { display: block; page-break-inside: avoid; } div.output { display: block; page-break-inside: avoid; } }
One by One [ 1 x 1 ] Convolution - counter-intuitively useful¶
March 25, 2016
Whenever I discuss or show GoogleNet architecture, one question always comes up - "Why 1x1 convolution ? Is it not redundant ?
Deep Learning software installation guide on fresh Ubuntu¶
January 15, 2016
Guide to building and installing CUDA, CuDNN, OpenCV, FFMPEG, Theano, Tensorflow, Keras, Lasagne, Torch and Caffe. It also includes common issues faced and recommended libraries and versions.
Extension of Hidden Markov Models – Formulas in two state HMM¶
February 10, 2015
Extending the hidden markov models, where the current state is affected by past two states. This can be useful in simulation of games or in less sophisticated pricing models (more accurate would be an exponential decay of all past terms).
Pseudo loss function in distributed Adaboost¶
December 10, 2014
This is the sketch of proof of correctness of , pseudo loss function, in the case of distributed boosting algorithm. It has been known that boosting of functions with each applicator with more than 0.5 accuracy is sufficient to guarantee lowest minimum accuracy in iterative pooling. A similar sketch is presented for the distributed case…
Dokuwiki for Personal Notebook and Note taking¶
March 8, 2013
As a researcher, you soon start wondering if you had centralized all your notes, possibly digitized them, life would be much better. Recently when I had to make a tough choice of leaving all my notes from years when I am about to shift country (due to limited air travel baggage), I wish I had them on computer. Since I will be making more…
Solutions to Hamilton-Jacobi-Bellman under uncertainity¶
March 2, 2013
After doing some reading on decision under un-certainity, I get the feeling that this I will be looking more into this. More so because I have the feeling like there is more to this field, lot of unknowns yet(which is still partly due to my lack of profound knowledge in the field). I feel this field is yet to mature.
Moments and transition probability of Trinomial Tree¶
January 30, 2013
I just switched from using WP-Latex to MathJax since Chrome is soon going to have built-in support for MathJax and it is easier to recover the tex codes for the reader.
Pascal’s Triangle in Standard ML¶
January 26, 2013
It has been a while that I posted something (grad school applications !). For past few weeks I have been learning Standard ML (SML), my first foray into functional programming language. I must say, I was skeptical at first due to ‘no-state’ concept but it is turning out to be great experience. Recursion can only be…
Monte Carlo Simulation in R¶
November 19, 2012
While the last post dealt with Bootstrap Sampling, no sampling discussion can be complete without discussion ‘Monte Carlo’ simulation. Readers please note, I will not discuss “MCMC (Markov Chain Monte Carlo)” (perhaps in the future). MCMC primarily deals with distribution of equilibrium of the given Markov Chain.
Bootstrap sampling in R¶
November 19, 2012
While there are lot of fancy and featured enabled simulation and sampling softwares, nothing is better than writing codes from groudup. While the sampling code given below is not much featured enabled but anyone with some background can easily add the features as they require. The bootstrapping could be used in several scenarios…
Notes on Discrete Painlevé Equations¶
November 10, 2012
Painlevé equations are used to solve non-linear second order differential equation. For more on Painleve’s work see this and for short refresher on Painlevé equations, Encylopedia of Mathmematics.
Hello ‘Lucene’ World – Web Implementation¶
October 20, 2012
Lucene is a Java Library by Apache used extensively for making custom search engines and indexing. Here are few of the features of Lucene, straight from the Lucene’s homepage :
Why Simulated Annealing works¶
October 7, 2012
Of all optimization methods, Simulated Annealing is one of the most fascinating one. If you need a quick refresher in Simulated Annealing then see this slide. Is Simulated Annealing better than other techniques in finding the global optima ? Perhaps. I will discuss why I think it is one of the best optimization technique and why so.
Value of Interdisciplinary Research¶
September 23, 2012
As the knowledge gets more and more advanced we find ourselves segmenting the branches of science into more specialized areas. This has lead to tunnel visioning of higher education. While this is not a bad thing, the consequences of it is when an intricate problem arises whose solution may be interdisciplinary, it remains unsolved or…
When in Pain: Read¶
September 13, 2012
Human life with all its color and variety has one thing in common, feeling of joy and sadness. No amount of wealth or luck can save you from either. Different people have different means to cope with tumultuous rides of emotion. While everyone has good handle on joyous days, the gloomy days elapse in a lot of pain.
Clipboard to Email – Python Code¶
August 25, 2012
Very often I have to send lot of stuff (links, texts, comments, notes) to a particular Email Address. Initially I started using www.emailtheweb.com , it has a nice bookmark button (when you press the same, the webpage in the background is emailed). This was not the swiftest method, since it had multiple clicks, few pages to visit, accept…
Protocol Engineering for Vehicular Communication – Falko Dressler¶
August 18, 2012
This is the notes from the lecture I attended by Dr Falko Dressler, Professor of Computer Science at University of Innsbruck.
Qubit operations and Quantum Measurement in Mathematica¶
August 11, 2012
This is a short note on how to get started with Quantum Computing with Mathematica.
How math can help discretionary traders¶
July 25, 2012
If you are doing discretionary trading you don’t want to trade on too many different instruments. It needs tremendous focus to concentrate on the flow of each tick. Choose a single instrument which has enough liquidity but is not among the most traded. Dangers of scalping say SPY or QQQQ or USDEUR is that they are traded by far too many…
The Birthday Paradox – Proof¶
July 21, 2012
Above statement is briefly The Birthday Paradox. If you need more information kindly see Short Version or the Long Version and if you prefer video then Birthday Paradox – Youtube
Paradigm towards shared Branch Banking¶
July 8, 2012
Long time ago when I was in my native town I had to do some banking operations on my account. To my surprise I found that the place didn’t have a branch of the bank where I had my account.