Is this you? As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.
Claim your profile
Articles by Andre Ye
[2311.02294] LLMs grasp morality in concept
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Recurrent Neural Networks, Explained and Visualized from the Ground Up Original
Unsplash Recurrent Neural Networks, Explained and Visualized from the Ground Up With an application to machine translation Recurrent Neural Networks (RNNs) are neural networks that can operate sequentially. Although they’re not as popular as they were even just several years ago, they represent an important development in the progression of deep learning and are a natural extension of…
Recurrent Neural Networks, Explained and Visualized from the Ground Up
The design of the Recurrent Neural Network (1985) is premised upon two observations about how an ideal model, such as a human reading text, would process sequential information: It should track the information 'learned' so far so it can relate new information to previously seen information. To understand the sentence "the quick brown fox jumped over the lazy dog", I need to keep track of the words 'quick' and 'brown' to understand later that these apply to the word 'fox'.
What Does It Really Mean for an Algorithm to ‘Learn’?
Two general perspectives and some psychology When one first encounters machine learning, one often rushes through algorithm after algorithm, technique after technique, equation after equation. But it is afterwards that one can reflect on the general trends across the knowledge that they have acquired. What it means to ‘learn’ is a very abstract concept. The goal of this article is to provide two general interpretations of what it means for a machine to learn.
An Equation for Intelligence
From Francois Chollet: “On the Measure of Intelligence” Throughout history, the measure of intelligence has been changed wildly to adapt to shifting thought and attitudes towards different groups of people and animals. In recent decades, the possibility of intelligence is being extended towards machines.
An Equation for Intelligence
Throughout history, the measure of intelligence has been changed wildly to adapt to shifting thought and attitudes towards different groups of people and animals. In recent decades, the possibility of intelligence is being extended towards machines.
Big Data Is Not the Way to Go
The significance of the data distribution A lot of discourse around deep learning would have you believe that more data is better. This is a fairly intuitive idea. Data represents the ‘real world’ in some capacity, and models training on such data learn a sampled representation of this ‘real world’. Ergo, the more sampled data, the closer the model’s understanding approximates the ‘real world’. This proposition appears to be corroborated by, well, the entire history of deep learning.
Big Data Is Not the Way to Go
By Andre Ye a la unsplash The significance of the data distribution A lot of discourse around deep learning would have you believe that more data is better. This is a fairly intuitive idea. Data represents the 'real world' in some capacity, and models training on such data learn a sampled representation of this 'real world'. Ergo, the more sampled data, the closer the model's understanding approximates the...
Deep Learning Is Not ‘Worse’ than Trees on Tabular Data
Moving beyond classical conceptions of what tabular data can be Deep learning has become the public and private face of Artificial Intelligence. When one talks casually about Artificial Intelligence with friends at a party, strangers on the street, and colleagues at work, it is almost always on the exciting models which generate language, create art, synthesize music, and so on. Massive and intricately designed deep learning models power most of these exciting machine capabilities.
Deep learning is not 'worse' than trees on tabular data
By Andre Ye A la pixabay. Moving beyond classical conceptions of what tabular data can be Deep learning has become the public and private face of Artificial Intelligence. When one talks casually about Artificial Intelligence with friends at a party, strangers on the street, and colleagues at work, it is almost always on the exciting models which...
What the Human Brain Has That Deep Learning Desperately Needs: A Guide to Zero-Shot Learning
Source: Unsplash And the Embarrassingly Simple ZSL Algorithm Deep learning has a big problem: it needs to gobble up massive quantities of data before it can generalize reasonably well and become useful. This is in fact one of the limitations of deep learning that restrict its applications in many fields where data is either not abundant or difficult to obtain.
Visualizing How Different ML Models Operate in the Feature Space
Intuitively Understand the Nature of Every Machine Learning Model I remember when I was learning machine learning models — KNN, Logistic Regression, SVM, Decision Tree, and so on — I learned only the theory. Because all these algorithms have lots of mathematics involved, most people only learn the shallow theory. However, often the understanding of a model is not deep enough to understand the nature of the algorithm.
4 Python AutoML Libraries Every Data Scientist Should Know
Source: Unsplash Automated Machine Learning, often abbreviated as AutoML, is an emerging field in which the process of building machine learning models to model data is automated. AutoML has the capability to make modelling easier and more accessible for everyone. If you’re interested in checking out AutoML, these four Python libraries are the way to go. A comparison will be provided at the end.
Meet Your New Best Exploratory Data Analysis Friend
Seaborn’s pairplot is magical: at its most simple, it gives us a rich and informational visual representation of univariate and bivariate relationships within the data. For instance, consider two pairplots below, created with one line of code, sns.pairplot(data) (the second adding hue=’species’ as a parameter).
5 Lesser-Known Seaborn Plots Most People Don’t Know
Seaborn is one of the most popular visualization libraries in Python and offers a vast array of plotting methods, some of which many are not familiar with. In this article, five relatively unknown Seaborn plots (and one Easter egg) will be introduced, along with diagrams, potential use cases, code, and explanations. Boxenplots Boxplots are notoriously bad visualizations because they hide distributions with a few potentially misleading statistical representations of the data.
Every ML Engineer Needs to Know Neural Network Interpretability
Neural Networks are the pinnacle of machine learning: they can model extremely complex functions by matching it with an equally complex structure. Full of nonlinearities and information propagation streams, the mindset with data scientists has often been “what we gain in power we sacrifice with interpretability.” Indeed — the black-box nature of the algorithm has been mostly accepted, but it shouldn’t be so.
5+ Simple One-Liners You’ve Been Looking For To Level Up Your Python Visualization
Line plot into area chart Consider the following standard line plot, created with seaborn’s lineplot, with the husl palette and whitegrid style. The data is generated as a sine wave with normally distributed data and elevated above the x-axis. With a few styling choices, the plot looks presentable. However, there is one issue: by default, Seaborn does not begin at a zero baseline, and the numerical impact of the y-axis is lost.
MissForest: The Best Missing Data Imputation Algorithm?
Missing data often plagues real-world datasets, and hence there is tremendous value in imputing, or filling in, the missing values. Unfortunately, standard ‘lazy’ imputation methods like simply using the column median or average don’t work well. On the other hand, KNN is a machine-learning based imputation algorithm that has seen success but requires tuning of the parameter k and additionally, is vulnerable to many of KNN’s weaknesses, like being sensitive to being outliers and noise.
Multicollinearity Impacts Your Data Science Project More Than You Know
Source: Unsplash Multicollinearity is likely far down on a mental list of things to check for, if it is on a list at all. This does, however, appear almost always in real-life datasets, and it’s important to be aware of how to address it. As its name suggests, multicollinearity is when two (or more) features are correlated with each other, or ‘collinear’. This occurs often in real datasets, since one measurement (e.g. family income) may be correlated with another (e.g. school performance).
Every Complex DataFrame Manipulation, Explained and Visualized Intuitively
Pandas offers a wide range of DataFrame manipulations, but many of them are complex and may not seem approachable. This article will present 8 essential DataFrame manipulation methods which cover almost all of the manipulation functions a data scientist would need to know. Each method will include an explanation, visualization, code, and tricks to remember it. All images by author.
The Dark Side of Transfer Learning
If some of the most key benchmark models in deep learning have been shown to be vulnerable to adversarial attacks, we should ask ourselves about the security implications for transfer learning models that fine-tune from those vulnerable benchmark models.
A Critical Look at How Data Science is “Taught”
Unsplash Although we may not think about it much, data science (and computer science broadly) is particularly interesting from an educational/pedagogical standpoint.
What They Don’t Tell You — 4 Ways Humans Still Vastly Outperform AI
ARTIFICIAL INTELLIGENCE Photo by Andre Hunter on Unsplash AI systems seem so intelligent because they give more exposure to achievements that reflect it. Reality tells us otherwise. Every time there’s a notable breakthrough in AI, we only hear how intelligent and skilled the systems are getting. In 2012 Hinton’s team got 63% top-1 accuracy on the ImageNet challenge. A few years later, a system topped human performance by achieving a striking +90% top-1 accuracy.
How Instacart Uses Data Science to Tackle Complex Business Problems
Photo by Scott Warman on Unsplash Instacart is the company that half of American households use to get groceries from their favorite grocery stores like Costco, Wegman’s, Whole Foods, Petco, and more without even leaving the house. Its unique business model makes it a fascinating case study on how data science can be used in the industry to tackle thorny business problems.
Using Mathematical Thinking to Find the Perfect Apartment
Photo by Chait Goli from Pexels After an intensive couple of months of interviewing, I landed an excellent data science role at an exciting company. I had gone back to school to do a Masters in Data Science in 2020 and, getting ready to settle in, I was now on the lookout for an awesome apartment. To help me find the best place in the shortest amount of time, I created a concise measuring tool for the apartments I’d be looking at and then planned to use that to find the best one.
Removing the “Adversarial” in Generative Adversarial Networks
Source Generative Adversarial Networks (GANs) are an important development in deep learning; its formulation inspired a new generation of model ensembles that interact with each other in ways to produce incredible results. Competition is in the blood of GAN design — “adversarial” is in its name. Two models, the discriminator and the generator, compete against each other.
The Evolution of Facial Recognition - A Case Study in the Transformation of Deep Learning
Machine learning has often been described as the study of “algorithms that create algorithms”. To a certain extent, this is true — machine learning finds the best model to fit the data. It is the process by which we attain the algorithm that can give predictions on the data. Yet, there is something slightly off about this description: it erases human task-specific involvement. In reality, humans have a heavy role in ensuring that machine learning algorithms work with the given data.
Deeper Neural Networks Lead to Simpler Embeddings
Recent research is increasingly investigating how neural networks, being as hyper-parametrized as they are, generalize. That is, according to traditional statistics, the more parameters, the more the model overfits. This notion is directly contradicted by a fundamental axiom of deep learning: Increased parametrization improves generalization. Although it may not be explicitly stated anywhere, it’s the intuition behind why researchers continue to push models larger to make them more powerful.
Can Transformers Solve This 90-Year-Old Classic Computer Science Problem Better Than Human Algorithms?
The Travelling Salesman Problem was formulated in 1930, and is a classical computer science problem for optimization. It’s a simple problem: Given a list of cities and the distances between each pair of cities, what is the shortest possible route that visits each city exactly once and returns to the origin city? For instance, we may be given distances between these four cities, as given below.
Machine Learning Goes Quantum: A Glance at an Exciting Paradigm Shift
Source. Revolutionizing traditional Transfer Learning, k-Means, and the CNN Quantum computing is a buzzword that’s been thrown around quite a bit. Unfortunately, despite its virality in pop culture and quasi-scientific Internet communities, its capabilities are still quite limited. As a very new field, quantum computing presents a complete paradigm shift to the traditional model of classical computing.
A Fascinating & State-of-the-Art Method to Train Neural Networks on Dishonest Data
Source: Unsplash Oh no! You find out that your data is corrupted — there enough instances of training data being attached to the incorrect label for it to be significant enough. What should you do? If you want to be a radical optimist, you could think of this data corruption as a form of regularization — depending on the level of corruption.
Can Machine Learning Ever Be “Fair” - and What Does That Even Mean?
Source: Unsplash Machine learning has begun to pervade all aspects of life — even those protected by anti-discrimination law. It’s being used in hiring, credit, criminal justice, advertising, education, and more. What makes this a particularly difficult problem, though, is that machine learning algorithms seem to be more fluid and intangible. Part of the design of complex machine learning algorithms is that they are difficult to interpret and regulate.
The Ultimate Scikit-Learn Machine Learning Cheatsheet
By Andre Ye, Cofounder at Critiq, Editor & Top Writer at Medium. Source: Pixabay. There are several areas of data mining and machine learning that will be covered in this cheat-sheet: Predictive Modelling. Regression and classification algorithms for supervised learning (prediction), metrics for evaluating model performance. Methods to group data without a label into clusters: K-Means, selecting cluster numbers based objective metrics. Dimensionality Reduction.
Artificial Intelligence (AI) - Top Use Cases and Technologies Used Today
Artificial intelligence is impacting the way in which enterprises engage with consumers. AI is no more science fiction and is already being adopted by public and private organisations globally.
Underspecification: The Dangerously Underdiscussed Problem Facing Machine Learning
good model loss ≠ good learning All machine learning models originate in the computer lab. They’re initialized, trained, tested, redesigned, trained again, fine-tuned, tested yet again before they are deployed. Afterwards, they fulfill their duty in epidemiological modelling, stock trading, shopping item recommendation, and cyber attack detection, among many other purposes.
Be Careful of This Data Science Mistake I Wasted 30 Hours Over
· Editors' PicksFeaturesExploreGrowContribute Source: Unsplash The model had been training across several sessions for many days on an image recognition competition. It was a relatively simple, and scored about a 0.9 AUC initially — the metric for the competition, which is between 0 and 1. I didn’t expect much from it at all.
6 Types of Neural Networks Every Data Scientist Must Know
Photo by Michael Dziedzic on Unsplash DEEP LEARNING, NEURAL NETWORKS The most common types of Neural Networks and their applications Neural networks are robust deep learning models capable of synthesizing large amounts of data in seconds. There are many different types of neural networks, and they help us in a variety of everyday tasks from recommending movies or music to helping us buy groceries online.
How We, Two Beginners, Placed in Kaggle Competition Top 4%
Source. If you’ve been keeping up with the Kaggle News, you may be familiar with the Mechanisms of Action competition by the Laboratory for Innovation Science at Harvard recently closed. I’m proud to say that my partner, Andy Wang, and I managed to place in the top 4% — 152nd out of 4,373 teams. What’s interesting, though, is that we’re relatively new to Kaggle competition.
A Beginner’s Guide to Decision Tree
Image on Google We make several decisions in life. Few of them are choosing to save or spend money, buying a home with a partner, breaking up with partner, getting a pet, choice of degree to study, choosing a University, quitting a job, quitting smoking, getting married, having children, buying a new car and many more.. Recent addition — Should I get a COVID test?
Turning Non-Image Data into Images for Classification is Surprisingly Effective
Study into convolutional neural networks (CNNs) has risen dramatically in recent years as the computing power and data is being increasingly made available. CNNs, however, are exclusively used for image data — and why not? They were designed for such a purpose. While they’re not perfect (in fact, far from it), CNNs have been extensively studied, such that they exist in wide shape and form.
BFS VS DFS
Back at again with the data structure and algorithm journal. In the last journal of the data structure, we talk about binary search trees. In this article, we are going to dive into how we traversal a tree. Not necessarily a binary search tree, it could be a binary tree that’s unsorted or a ternary tree or just any other tree that has no particular order. Let’s contrast it with a linked list to traverse a tree for every node one time.
Data Analysis - Time Series, Resample in Pandas
We’ll be only using pandas and NumPy. import pandas as pd import numpy as np Let’s generate some random data first. # Create a 2-month range datatime datetime = pd.date_range(start='01-01-2020',end='02-28-2020',freq='1min')# Generate a randome integer array between 0 and 100 as the value.
Every Complex DataFrame Manipulation, Explained & Visualized Intuitively
By Andre Ye, Cofounder at Critiq, Editor & Top Writer at Medium. Pandas offers a wide range of DataFrame manipulations, but many of them are complex and may not seem approachable. This article presents 8 essential DataFrame manipulation methods that cover almost all of the manipulation functions a data scientist would need to know. Each method will include an explanation, visualization, code, and tricks to remember it.
Markov Decision Process (MDP): The Father of Reinforcement Learning
Understanding Markov Decision Process (MDP) and How It Shaped Reinforcement Learning Photo by Clarisse Croset on Unsplash Lesson Learned From Second Meeting of AWS x JML DeepRacer Bootcamp Big Thanks Once Again to AWS and Jakarta Machine Learning !! In the first part of this series, we’ve learned about the basic concept of Reinforcement Learning (RL) and how it works inside the autonomous racing car.
Natural Language Processing: An Analysis of Sentiment.
In this project I will investigate the basics of Natural Language Processing (NLP) and aim to predict whether a sample of tweets are either positive or negative. This will consist of combining machine learning principles with text. I will use mathematics and statistics to get the text in a format that algorithms can understand.
The Two-Headed Neural Network Shaking Up Image Recognition
The Magic of Siamese Networks Deep neural networks have a big problem — they’re constantly hungry for data. When there is too little data — an amount that would be acceptable for other algorithms — deep neural networks have tremendous difficulty generalizing. This phenomenon highlights a gap between human and machine cognition; humans can learn complex patterns with few training examples (albeit at a slower rate).
XGBoost, LightGBM, and Other Kaggle Competition Favorites
An Intuitive Explanation and Exploration Kaggle is the data scientist’s go-to place for datasets, discussions, and perhaps most famously, competitions with prizes of tens of thousands of dollars to build the best model. With all the flurried research and hype around deep learning, one would expect neural network solutions to dominate the leaderboards.
What Happens When Neural Networks Hallucinate
Source: Google AI Blog. Image free to share. the mechanics & implications of deep dreaming What happens to a human mind when it is hallucinating? Most narcotics that produce hallucinations — experiences in which you see, hear, feel, or smell something that does not exist — do so by shooting the brain with ‘pleasure chemicals’.
How to Systematically Fool an Image Recognition Neural Network
and why it matters… a lot Convolutional neural networks — CNNs — form the basis for image recognition, which is undoubtedly one of the most important applications of deep learning. Unfortunately, much of research in deep learning is done in the ‘perfect-world’ constraints of datasets in pursuit of a few percentage points in accuracy. Thus, we’ve developed architectures that work tremendously well in theoretical tests but not necessarily so in the real world.
Monte Carlo Methods, Made Simple
Using Chaos to Find Clarity Imagine a 10 by 10 square on a coordinate grid. Some shape is drawn on that grid, but you don’t know what it looks like. However, you can query a function f(x, y) where (x, y) is the coordinate and the output is either 1 (in the shape) or 0 (not in the shape). How would you find the area of the shape? The answer is quite simple.
Show More
loading
Actions
Is this you?
As a journalist, you can create a free Muck Rack account to customize your profile, list your contact preferences, and upload a portfolio of your best work.Get in touch with Andre
Contact Andre, search articles and posts on X, monitor coverage, and track replies from one place.
Learn more about Muck Rack