About
As SVP of Digital Transformation at Fiserv, I focus on digital transformation and…
Articles by Tom
-
Offer of pro-bono assistance to COVID-19 Researchers
Offer of pro-bono assistance to COVID-19 Researchers
By Tom Eck
-
10 secrets for sabotaging a micro-services project
10 secrets for sabotaging a micro-services project
By Tom Eck
Contributions
-
What are the best practices for improving deep learning interpretability with regularization?
I fail to see the connection between regularization and model interpretability. Regularization is a set of techniques to improve the quality and possibly speed of training a model. This is a training issue Interpretability is the tough task of interpreting WHY the model produced the response it gave This is an inference issue
-
How can you measure the impact of a performance issue?
It comes down to the impact felt by the users of the application. Were they unable to perform the task they depended on the app for? Was there a loss of revenue? How long did the disruption last Also, there’s potential impact to the brand reputation of the company offering that product. And in my industry (financial services) there can be monetary penalties as well
-
What are the most important AI skills for professionals?
Python has ruled the world of ML and AI for years. However it is not compiled, so modern frameworks often sit atop C/C++ libraries to accelerate certain operations (eg matrix manipulation) Elon Musk, however, is stating that Rust will be the king of languages for AI going forward. While Rust is indeed highly performant, I find it obscure and is not the right approach for doing quick and dirty experimentation.
-
What are the best ways to incorporate AI into statistical programming?
Flipping the question around, AI (at least neural-nets) is essentially 3 mathematical disciplines: - probability/statistics (generating a prob distribution across all potential outputs) - linear algebra (matrix operations) - calculus (back-prop=chain rule)
Activity
-
🐳 Big news! Hugging Face and Google Cloud have joined forces to bring you a collection of Deep Learning Containers (DLCs) to transform the way you…
🐳 Big news! Hugging Face and Google Cloud have joined forces to bring you a collection of Deep Learning Containers (DLCs) to transform the way you…
Liked by Tom Eck
-
If you're working on a document AI use case where layouts can get pretty complex, I recommend leveraging Google's Gemini models with few-shot…
If you're working on a document AI use case where layouts can get pretty complex, I recommend leveraging Google's Gemini models with few-shot…
Liked by Tom Eck
-
What's really going on in machine learning? Just finished a deep dive using (new) minimal models. Seems like ML is basically about fitting together…
What's really going on in machine learning? Just finished a deep dive using (new) minimal models. Seems like ML is basically about fitting together…
Liked by Tom Eck
Experience
Education
-
University of Pennsylvania
-
Activities and Societies: Received $50,000 award to develop my software which identified genetic aberrations linked to specific subclasses of cancer
Final GPA 4.0
-
-
GPA 4.0
I am studying how to apply genomics and cheminformatics to the design of new drugs. -
-
Final GPA 4.0
Thesis project involved developing artificial intelligence software (ART 2 neural network) to identify human voices from very short speech samples. This work was applied and utilized at the Department of Defense, US Army, Electronic Warfare -
-
Activities and Societies: Rugby
Pre-Med
Computer Science minor
Licenses & Certifications
Volunteer Experience
-
Founder of Technology and Coding Club
Christian Brothers Academy
- Present 8 years 6 months
Education
I founded the technology club with my son at his High School so that I can share my knowledge and enthusiasm for programming and building tech solutions.
-
Youth Group Leader
Numerous churches
- 17 years
Children
Energetic youth group leader who ran meetings, chaperoned many retreats, and led youth group rock bands.
-
Co-founder and inaugural Coach
Wall Township Girls Lacrosse Club
- 3 years 1 month
Children
Publications
-
Detecting Allosteric Sites of HIV-1 Reverse Transcriptase by X-ray Crystallographic Fragment Screening
Journal of Medicinal Chemistry
HIV-1 reverse transcriptase (RT) undergoes a series of conformational changes during viral replication and is a central target for antiretroviral therapy. The intrinsic flexibility of RT can provide novel allosteric sites for inhibition. Crystals of RT that diffract X-rays to better than 2 Å resolution facilitated the probing of RT for new druggable sites using fragment screening by X-ray crystallography. A total of 775 fragments were grouped into 143 cocktails, which were soaked into crystals…
HIV-1 reverse transcriptase (RT) undergoes a series of conformational changes during viral replication and is a central target for antiretroviral therapy. The intrinsic flexibility of RT can provide novel allosteric sites for inhibition. Crystals of RT that diffract X-rays to better than 2 Å resolution facilitated the probing of RT for new druggable sites using fragment screening by X-ray crystallography. A total of 775 fragments were grouped into 143 cocktails, which were soaked into crystals of RT in complex with the non-nucleoside drug rilpivirine (TMC278). Seven new sites were discovered, including the Incoming Nucleotide Binding, Knuckles, NNRTI Adjacent, and 399 sites, located in the polymerase region of RT, and the 428, RNase H Primer Grip Adjacent, and 507 sites, located in the RNase H region. Three of these sites (Knuckles, NNRTI Adjacent, and Incoming Nucleotide Binding) are inhibitory and provide opportunities for discovery of new anti-AIDS drugs.
Other authorsSee publication -
Preneoplastic lesion growth driven by the death of adjacent normal stem cells
Proceedings of the National Academy of Sciences
Clonal expansion of premalignant lesions is an important step in the progression to cancer. This process is commonly considered to be a consequence of sustaining a proliferative mutation. Here, we investigate whether the growth trajectory of clones can be better described by a model in which clone growth does not depend on a proliferative advantage. We developed a simple computer model of clonal expansion in an epithelium in which mutant clones can only colonize space left unoccupied by the…
Clonal expansion of premalignant lesions is an important step in the progression to cancer. This process is commonly considered to be a consequence of sustaining a proliferative mutation. Here, we investigate whether the growth trajectory of clones can be better described by a model in which clone growth does not depend on a proliferative advantage. We developed a simple computer model of clonal expansion in an epithelium in which mutant clones can only colonize space left unoccupied by the death of adjacent normal stem cells. In this model, competition for space occurs along the frontier between mutant and normal territories, and both the shapes and the growth rates of lesions are governed by the differences between mutant and normal cells' replication or apoptosis rates. The behavior of this model of clonal expansion along a mutant clone's frontier, when apoptosis of both normal and mutant cells is included, matches the growth of UVB-induced p53-mutant clones in mouse dorsal epidermis better than a standard exponential growth model that does not include tissue architecture. The model predicts precancer cell mutation and death rates that agree with biological observations. These results support the hypothesis that clonal expansion of premalignant lesions can be driven by agents, such as ionizing or nonionizing radiation, that cause cell killing but do not directly stimulate cell replication.
Other authorsSee publication -
STAC: A method for testing the significance of DNA copy number aberrations across multiple array-CGH experiments
Journal of Genome Research
Regions of gain and loss of genomic DNA occur in many cancers and can drive the genesis and progression of disease. These copy number aberrations (CNAs) can be detected at high resolution by using microarray-based techniques. However, robust statistical approaches are needed to identify nonrandom gains and losses across multiple experiments/samples. We have developed a method called Significance Testing for Aberrant Copy number (STAC) to address this need.
-
Identification of common deletions in breast cancers from non-BRCA1/BRCA2 families using array-based comparative genomic hybridization
Cancer Research
Germline mutations in BRCA1 and BRCA2 account for 20-40% of large multiple case breast cancer families. However, breast cancer susceptibility in the majority of other families (termed BRCAX families) is not due to BRCA1 and BRCA2 mutations. Here we applied array-based Comparative Genomic Hybridization (aCGH) to characterize copy number alterations in 44 breast cancers from the 38 BRCAX families currently included in an international collaborative linkage analysis.
-
An automatic text-free speaker recognition system based on an enhanced Art 2 neural architecture
Information Sciences
The adaptive-resonance-theory (ART) architectures comprise a family of neural networks that efficiently self-organize stable recognition codes in response to arbitrary sequences of input patterns. This paper focuses on the ART 2 network architecture that is designed for the processing of both binary- and analog-valued patterns. The problems encountered in implementing the primary ART 2 architecture as originally presented by Gail Carpenter and Stephen Grossberg are discussed. An enhanced ART 2…
The adaptive-resonance-theory (ART) architectures comprise a family of neural networks that efficiently self-organize stable recognition codes in response to arbitrary sequences of input patterns. This paper focuses on the ART 2 network architecture that is designed for the processing of both binary- and analog-valued patterns. The problems encountered in implementing the primary ART 2 architecture as originally presented by Gail Carpenter and Stephen Grossberg are discussed. An enhanced ART 2 architecture that receives its input from a functional-link preprocessor is proposed. Experimental results that demonstrate this architecture's superior performance over the previous ART 2 architecture are provided. Finally, a text-free speaker recognition system that employs the enhanced ART 2 architecture as its classifier is described.
Other authors -
Recommendations received
5 people have recommended Tom
Join now to viewMore activity by Tom
-
I’m seeing increasing interest in running #AI models on-premise. Here are three primary considerations off the top of my head: 1️⃣ Cost 2️⃣…
I’m seeing increasing interest in running #AI models on-premise. Here are three primary considerations off the top of my head: 1️⃣ Cost 2️⃣…
Posted by Tom Eck
-
The Nvidia Pruning and Distillation paper is a technical masterpiece. LLMs like Llama 3.1 405B and NVIDIA Nemotron-4 340B excel in many tasks but…
The Nvidia Pruning and Distillation paper is a technical masterpiece. LLMs like Llama 3.1 405B and NVIDIA Nemotron-4 340B excel in many tasks but…
Liked by Tom Eck
-
Free Hands-On Hugging Face Tutorial | Transformers, AI Pipeline, Fine Tuning LLM, GPT, Sentiment Analysis 🎥Watch this 15-minute video to jumpstart…
Free Hands-On Hugging Face Tutorial | Transformers, AI Pipeline, Fine Tuning LLM, GPT, Sentiment Analysis 🎥Watch this 15-minute video to jumpstart…
Liked by Tom Eck
-
This article should be used as a cautionary tale to insure that you have the proper controls around the data sources that #GenAI and associated…
This article should be used as a cautionary tale to insure that you have the proper controls around the data sources that #GenAI and associated…
Shared by Tom Eck
-
The ability of LLMs to generate code seems magical, but consider the volume of previously-existing code that they are trained on. "Back in the Day"…
The ability of LLMs to generate code seems magical, but consider the volume of previously-existing code that they are trained on. "Back in the Day"…
Shared by Tom Eck
-
🔬 🤓 Scientists, researchers, nerds in general: If you are working on something cool, but not quite ready to quit your job and start a company, we…
🔬 🤓 Scientists, researchers, nerds in general: If you are working on something cool, but not quite ready to quit your job and start a company, we…
Liked by Tom Eck
-
Want to combine quantization and benefit from `torch.compile()` for diffusion models? Enter `torchao`! It allows us to run Flux under 17GBs without…
Want to combine quantization and benefit from `torch.compile()` for diffusion models? Enter `torchao`! It allows us to run Flux under 17GBs without…
Liked by Tom Eck
-
Wondering which ligand is more selective? Swap them! Solmaz Azimi and I show that alchemical receptor hopping and swapping are powerful…
Wondering which ligand is more selective? Swap them! Solmaz Azimi and I show that alchemical receptor hopping and swapping are powerful…
Liked by Tom Eck
-
AI and Productivity…a quick listen with Malcolm Gladwell, who is the best interviewer I’ve met.
AI and Productivity…a quick listen with Malcolm Gladwell, who is the best interviewer I’ve met.
Liked by Tom Eck
-
Fantastic illustration of the ever-increasing performance of LLM's. Two points: 1. OpenSource models are exerting pressure on closed-source to…
Fantastic illustration of the ever-increasing performance of LLM's. Two points: 1. OpenSource models are exerting pressure on closed-source to…
Shared by Tom Eck
-
This past spring, I fulfilled my lifelong goal of graduating from the University of Notre Dame... for a second time! After spending yet another…
This past spring, I fulfilled my lifelong goal of graduating from the University of Notre Dame... for a second time! After spending yet another…
Liked by Tom Eck
-
Wave hello to our rehab team aboard Big Blue as they head to shower at the MUSC Wellness Center! #StormDebby #TheATeam
Wave hello to our rehab team aboard Big Blue as they head to shower at the MUSC Wellness Center! #StormDebby #TheATeam
Liked by Tom Eck
-
💥 I’m on the lookout 🕶 for a Head of Digital & Media for Fiserv. But this isn’t just any leadership role. We’re seeking a visionary who can both…
💥 I’m on the lookout 🕶 for a Head of Digital & Media for Fiserv. But this isn’t just any leadership role. We’re seeking a visionary who can both…
Liked by Tom Eck
Other similar profiles
Explore collaborative articles
We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.
Explore MoreOthers named Tom Eck in United States
-
Tom Eck
-
Tom Eck
Portfolio Manager at Cortina Asset Management
-
Tom Eck
-
Tom Bombeck
Medical Device | Marketing | Business Development | Strategy
155 others named Tom Eck in United States are on LinkedIn
See others named Tom Eck