ProGAN: How NVIDIA Generated Images of Unprecedented Quality

Progressively growing GANs enables them to get bigger and more stable The people in the high resolution images above may look real, but they are actually not — they were synthesized by a ProGAN trained on millions of celebrity images. “ProGAN” is the colloquial term for a type of generative adversarial network that was pioneered at NVIDIA. It … Read more

Ways to Improve a Map Visualization

How to take a map visualization to the next level. First, I will cover two reasons why visualizing data using maps is often compelling to an audience. Then, I will cover three tips that will help you make the transition from good to exceptional when building map visualizations. Why Use a Map for a Data Visualization? … Read more

Robots that Reason

Inorganic knowledge traditions with model-based reinforcement learning This essay explores the concept of inorganic knowledge traditions capable of sequential improvement using model based reinforcement learning Many behavioral economists presently believe that there are two primary methods used by humans for strategic decision making. One is fast, intuitive and unconscious — what has been called System 1 thinking. … Read more

Regression Analysis: Linear Regression

3. Model Building in R I have used the dataset which contains the details of 2,201 flights. The descriptions of each variable are as below. 3.1) Datasets schedtime : the scheduled time of departure (using the 24-hour clock) carrier : the two-letter code indicating which airline operated the flight deptime : the actual departure time dest : the three-letter code … Read more

What’s the fuss about Regularization?

As a newbie to machine learning most people get excited when their training error starts reducing. They try hard further and it starts reducing even further, their excitement knows no bounds. They show their results to master Oogway ( elderly wise tortoise in Kungfu Panda) and he calmly says well not a good model you … Read more

Finding Local Events Using Twitter Data

Project by David Chen, Ashwin Gupta, Shruthi Krish, Raghav Prakash, Wei Wang Twitter is a social media platform that millions of users use to share updates about their lives. Often, these tweets are about local events happening around the user. Though news agencies report on local events, the time it takes an agency to learn … Read more

Predicting hospital length-of-stay at time of admission

Exploring an important healthcare performance metric Photo by Hush Naidoo on Unsplash Project Overview Predictive analytics is an increasingly important tool in the healthcare field since modern machine learning (ML) methods can use large amounts of available data to predict individual outcomes for patients. For example, ML predictions can help healthcare providers determine likelihoods of disease, … Read more

Using Markov Chain Monte Carlo method for project estimation

Using TensorFlow probability for Hamiltonian Sampling Free photo from One type of criticism I received for the previous work on project estimation is that the log-Normal distribution has short tails. And this is true, despite all the benefits of log-Normal distribution. The reason is very simple: when fitting the data to the distribution shape … Read more

Law and Word Order: NLP in Legal Tech

The law has language at its heart, so it’s not surprising that software that operates on natural language has played a role in some areas of the legal profession for a long time. But the last few years have seen an increased interest in applying modern techniques to a wider range of problems, so this … Read more

Keras with R: Predicting car sales

Keras is an API used for running high-level neural networks. The model runs on top of TensorFlow, and was developed by Google. In this particular example, a neural network will be built in Keras to solve a regression problem, i.e. one where our dependent variable (y) is in interval format and we are trying to … Read more

Modeling tree height and basal area in the Finger Lakes National Forest, NY

I tried my hand at using the R package, randomForest to create two regression models for tree height and basal area based off some lidar and field-collected data in the Finger Lakes National Forest, NY. Disclaimer: this project was my first real taste of R. Earlier in the semester I had done some simple learning into … Read more

Text Generation Using RNNs

Generate characters from Alice in Wonderland Introduction Text generation is a popular problem in Data Science and Machine Learning, and it is a suitable task for Recurrent Neural Nets. This report uses TensorFlow to build an RNN text generator and builds a high-level API in Python3. The report is inspired by @karpathy ( min-char-rnn) and … Read more

Introduction to Interactive Time Series Visualizations with Plotly in Python

Introduction to Plotly Plotly is a company that makes visualization tools including a Python API library. (Plotly also makes Dash, a framework for building interactive web-based applications with Python code). For this article, we’ll stick to working with the plotly Python library in a Jupyter Notebook and touching up images in the online plotly editor. When … Read more

Applying GANs to Super Resolution

SRGAN Results from Ledig et al. [3] Generative adversarial networks (GANs) have found many applications in Deep Learning. One interesting problem that can be better solved using GANs is super-resolution. Super-resolution is a task concerned with upscaling images from low-resolution sizes such as 90 x 90, into high-resolution sizes such as 360 x 360. In this … Read more

A Comprehensive Guide to Convolutional Neural Networks — the ELI5 way

Artificial Intelligence has been witnessing a monumental growth in bridging the gap between the capabilities of humans and machines. Researchers and enthusiasts alike, work on numerous aspects of the field to make amazing things happen. One of many such areas is the domain of Computer Vision. The agenda for this field is to enable machines … Read more

Time-series Forecasting Flow

A brief introduction on critical steps in demand forecasting Collecting The key here is the format of data storage. Intuitively we think of time-series data in the format below, also known as the wide format. However, the wide format is bad for SQL-based storage, as when we add new dates, we need to add another … Read more

Google Landmark Recognition using Transfer Learning

Image classification with 15k classes! Project by Catherine McNabb, Anuraag Mohile, Avani Sharma, Evan David, Anisha Garg Dealing with a large number of classes with very few images in many classes is what makes this task really challenging! The problem comes from a famous Kaggle competition, the Google Landmark Recognition Challenge. Training set contains over 1.2 … Read more

Part 3: Implementation in Java

Dec 14, 2018 In this article you will see how the theories presented in previous two articles can be implemented in easy to understand java code. The full neural network implementation can be downloaded, inspected in detail, built upon and experimented with. This is the third part in a series of articles: I assume you … Read more

Anime Recommendation engine: From Matrix Factorization to Learning-to-rank

Anime Obsession gone too far!! OtakusHenry Chang, Joey Chen, Guanhua Zhang, Preetika Srivastava and Cherry Agarwal The vast amount of data that is hosted on the internet today has led to the information overflow and thus there is a constant need to improve the user experience. A recommendation engine is a system that helps support … Read more

Home Remodeling Analysis Turned Excel Data Handling in Python

Why cleaning data is the most important step Original Project Mission: Find interesting insights to see where the remodeling market is headed Project Mission (Twist): How to handle well manicured excel data in Python (spoiler: neat is a deceptive word) Timeline: One week (I tell you, it’s not enough!) Project Findings for the Original Goal : … Read more

Finding Trends in NY Times Best Sellers

Every author dreams of writing full time, but the sad truth is that the majority of authors don’t make nearly enough to support themselves, let alone a family. If you hit the NY Times Best Seller list, your chances of making writing your career will be much higher, especially if it stays on the list … Read more

The complete guide for topics extraction with LDA (Latent Dirichlet Allocation) in Python

A recurring subject in NLP is to understand large corpus of texts through topic extraction. Whether you analyze users’ online reviews, product descriptions, or text entered in search bars, understanding key topics will always come in handy. Popular picture explaining LDA Before going into the LDA method, let me remind you that not reinventing the … Read more

Exploratory Data Analysis, Feature Engineering and Modelling using Supermarket Sales Data. Part 1.

The first thing I like to do when doing EDA on a dataset with a reasonable amount of numeric columns, is to check the relationship between my target variable and these numeric features. One quick way to do this is to use the seaborn heatmap plot. This seaborn heatmap takes the correlation matrix calculated on … Read more

Want to Cluster Text? Try Custom Word-Embeddings!

Tf-idf vectors with word-embeddings are analyzed for clustering effectiveness. The text corpus examples considered here indicate that custom word-embeddings can help improve clusterability of the corpus That is welcome news after our ho-hum results for text classification when using word-embeddings. In the context of classification we concluded that keeping it simple with naive bayes and tf-idf … Read more

Weekly Selection — Dec 14, 2018

3rd Wave Data Visualization By Elijah Meeks — 12 min read Imagine what it was like to do data visualization 30 years ago. It’s 1988 and you’re using Excel 2.0 for simple charts like pie charts and line charts, or maybe something like SPSS for more complicated exploration and Arc/Info for geospatial data visualization. Favorite

CSV Analysis with Amazon Athena

Executing standard SQL queries on your Amazon S3 bucket files Dec 14, 2018 “What’s Amazon Athena?”, I hear you ask. Good question. It’s one of Amazon Web Services’ amenities for architecture in the cloud. More specifically, Athena allows us to query data we hold in another service called Amazon Simple Storage Service (S3) using standard SQL … Read more

Using Analysis of Variance with Experimentation Data

“The one-way analysis of variance (ANOVA) is used to determine whether there are any statistically significant differences between the means of two or more independent (unrelated) groups (although you tend to only see it used when there are a minimum of three, rather than two groups)”. Having entered the world of digital analytics from a … Read more

A map for Machine Learning on AWS

Dec 14, 2018 It looks like Christmas is a little early this year 😉 Here’s a little something from me to all of you out there: a map to navigate ML services on AWS. With all the new stuff launched at re:Invent, I’m quite sure it will come in handy! This is very much work in … Read more

Get a glimpse of future using time series forecasting using Auto-ARIMA and Artificial Intelligence

Time Series Forecasting using Auto-ARIMA in python. AI and future Currently, there is a lot of development going on in Artificial intelligence research to get an accurate glimpse of the future. If any mathematical model predicts future data taking input as only time then that terminology called as time series forecasting. There are many machine learning and … Read more

The Naive Bayes Classifier

Joseph Catanzarite The Naïve Bayes Classifier is perhaps the simplest machine learning classifier to build, train, and predict with. This post will show how and why it works. Part 1 reveals that the much-celebrated Bayes Rule is just a simple statement about joint and conditional probabilities. But its blandness belies astonishing power, as we’ll see … Read more

Unpacking (**PCA)

Alright, better to implement PCA to get the image. Let’s start by making 5 *10 matrix, and take steps of the process. Matrix X import numpy as npX = np.random.rand(5,10) The column are variables (characteristics) and the row are samples(say, ‘cat’ or ‘dog’). What we want to do with this matrix is to get eigenvalues … Read more

Kuzushiji-MNIST – Japanese Literature Alternative Dataset for Deep Learning Tasks

Plus our VGG-ResNet ensemble model with state-of-the-art results MNIST, a dataset with 70,000 labeled images of handwritten digits, has been one of the most popular datasets for image processing and classification for over twenty years. Despite its popularity, contemporary deep learning algorithms handle it easily, often surpassing an accuracy result of 99.5%. A new paper … Read more

Closing the Sale: Predicting Home Prices via Linear Regression

Imports, Data Cleansing, and EDA Cleaning and EDA are important for this challenge as this data set contains many ordinal / categorical features that may be important in categorization and will need to be converted to numerical values. As a baseline, I imported the following libraries to clean, explore and model the training data. One of … Read more

Nobody UNDERSTANDS Me … But Soon, Artificial Intelligence Just Might

Our faces and voices can be analyzed for emotion. As I mentioned, biomimicry, or imitating natural design in the things we create, is critical in recreating this human tendency in AI. Our end goal is artificial empathy, which (for now, at least) describes a machine’s ability to recognize and respond to human emotion. Going in line … Read more

An introduction to high-dimensional hyper-parameter tuning

If you ever struggled with tuning Machine Learning (ML) models, you are reading the right piece. Hyper-parameter tuning refers to the problem of finding an optimal set of parameter values for a learning algorithm. Usually, the process of choosing these values is a time-consuming task. Even for simple algorithms like Linear Regression, finding the best … Read more

Preprocessing with sklearn: a complete and comprehensive guide

For aspiring data scientist it might sometimes be difficult to find their way through the forest of preprocessing techniques. Sklearn its preprocessing library forms a solid foundation to guide you through this important task in the data science pipeline. Although Sklearn a has pretty solid documentation, it often misses streamline and intuition between different concepts. … Read more

IcoOmen: Using Machine Learning to Predict ICO Prices

Methodology Choose inputs and outputs. Collect and aggregate the data. Prepare the data. Explore and attempt to understand the data. Choose a Machine Learning Model. Measure the performance of the Model. Save the Model. Use the Model to make predictions. 1. Choosing Inputs and Outputs Inputs Choosing the right inputs and outputs (in the case of … Read more

How to Predict Severe Traffic Jams with Python and Recurrent Neural Networks?

An Application of Sequence Model to Mine Waze Open Data of Traffic Incidents, using Python and Keras. In this tutorial, I will show you how to use RNN deep learning model to find patterns from Waze Traffic Open Data of Incidents Report, and predict if severe traffic jams will happen shortly. Interventions can be taken out … Read more

How to get the most out of Towards Data Science?

Our Readers’ Guide We have received feedback that some of you find it difficult to efficiently navigate our Medium publication. So we have put together a few bullet points that will hopefully aid your experience on our blog. Subscribe to our publication to receive our Monthly Edition and Weekly Selection directly in your mailbox. Follow us … Read more

Vaex: Out of Core Dataframes for Python and Fast Visualization

So… no pandas ?? There are some issues with pandas that the original author Wes McKinney outlines in his insightful blogpost: “Apache Arrow and the “10 Things I Hate About pandas”. Many of these issues will be tackled in the next version of pandas (pandas2?), building on top of Apache Arrow and other libraries. Vaex starts … Read more

Music Genre Classification with Python

Objective Companies nowadays use music classification, either to be able to place recommendations to their customers (such as Spotify, Soundcloud) or simply as a product (for example Shazam). Determining music genres is the first step in that direction. Machine Learning techniques have proved to be quite successful in extracting trends and patterns from the large … Read more

Named Entity Recognition (NER), Meeting Industry’s Requirement by Applying state-of-the-art Deep…

we are going to have a quick look at the architecture of four different state-of-the-art approaches by referring to the actual research paper and then we will move on to implement the one with the highest accuracy. Bidirectional LSTM-CRF: More details and implementation in keras. from the paper(Bidirectional LSTM-CRF Models for Sequence Tagging) 2. Bidirectional LSTM-CNNs: … Read more

The Importance of Being Recurrent for Modeling Hierarchical Structure

RNNs have inherent performance limitations For a while, it seemed that RNN’s were taking the Natural Language Processing (NLP) world by storm (from about 2014–17). However, we’ve recently started realizing the limitations of RNN’s, primarily that they are “inefficient and not scalable”. While there is great promise in overcoming these limitations by using more specialized … Read more