Case Study Example 1: An eCommerce Company Evaluation
You’re a Data Scientist / Business Analyst working for a new eCommerce company called A&B Co. (similar to Amazon) and you’ve been asked to prepare a presentation for the Vice President of Sales and the Vice President of Operations that summarizes sales and operations thus far. The summary should include (at a minimum) a summary of current state the business, current customer satisfaction, and a proposal of 2-3 areas where the company can improve.
Decision Theory
Decision theory and artificial intelligence typically try to compute something resembling argmax f(a). I.e., maximize some function of the action. This tends to assume that we can detangle things enough to see outcomes as a function of actions. For example, AIXI represents the agent and the environment as separate units which interact over time through clearly defined i/o channels, so that it can then choose actions maximizing reward.
More Robust Monotonic Binning Based on Isotonic Regression
Interactive Graphics with R Shiny
Well, R is definitively here to stay and made its way into the data science tool zoo. For me as a statistician, I often feel alienated surrounded by these animals, but R is still also the statistician’s tool of choice (yes, it has come to age, but where are the predators ..?) What was usually a big problem for us statistician, was to get our methods and models out to our customers, who (usually) don’t speak R. At this point Shiny comes in handy and offers a whole suite of bread and butter interface widgets, which can be deployed to web-pages and wired to R functions via all kinds of callback-routines.
Choosing hyper-parameters in penalized regression
In this post, I’m evaluating some ways of choosing hyper-parameters (a a and ? ? ) in penalized linear regression. The same principles can be applied to other types of penalized regresions (e.g. logistic).
Humanizing Customer Complaints using NLP Algorithms
Last Christmas, I went through the most frustrating experience as a consumer. I was doing some last minute holiday shopping and after standing in a long line, I finally reached the blessed register only to find out that my debit card was blocked. I could sense the old lady at the register judging me with her narrowed eyes. Feeling thoroughly embarrassed, I called my bank right away. To my horror, they told me that my savings account was hacked and thousands of dollars were already gone!
Mitigating algorithmic bias in predictive justice: 4 design principles for AI fairness
When Netflix gets a movie recommendation wrong, you’d probably think that it’s not a big deal. Likewise, when your favourite sneakers don’t make it into Amazon’s list of recommended products, it’s probably not the end of the world. But when an algorithm assigns you a threat score from 1 to 500 that is used to rule on jail time, you might have some concerns about this use of predictive analytics. Artificial intelligence (AI) has now permeated almost every aspect of our lives. Naturally, machine predictions cannot always be 100% accurate. But the cost of error dramatically increases when AI is implemented in high-stakes settings. This could include medicine to recommend new cancer treatments, or criminal justice to help judges assess a suspect’s likelihood of reoffending. In fact, one of the most controversial uses of AI in recent years has been predictive policing.
Dimensionality Reduction: ways and intuitions
After Big data applications became pervasive, the curse of dimensionality turns out to be more serious than expected. As a result, visualization and analysis became harder for this high dimensional space. Moreover, our insights and intuitions coming from the 2D and 3D world (e.g., Distances) no longer hold there. Another problem is that predictive models are more likely to have high variance when the number of predictors is high, which makes the models prone to overfitting. ?This article is about Dimensionality reduction techniques, that I learnt throughout my first semester as a Data Science Master’s student. For each technique, I will try to deliver the intuition, criteria and use-cases, and support the illustration in Python. For simplicity, I won’t go into the underlying mathematics of these techniques, but I will go through the questions that came to my mind when I studied them for the first time.
Plant AI – Deploying Deep Learning Models
So in my last post, I talked about how I built Plant AI?-?a Plant Disease detection model using Convolutional Neural Network. At the end, we had a model which we would be deploying in this post. The code for the Plant AI can be found here and the output here.
10 Gradient Descent Optimisation Algorithms
Gradient descent is an optimisation method for finding the minimum of a function. It is commonly used in deep learning models to update the weights of the neural network through backpropagation. In this post, I will summarise the common gradient descent optimisation algorithms that are used in popular deep learning frameworks (e.g. TensorFlow, Keras, PyTorch, Caffe). The purpose of this post is to make it easy to read and digest since there aren’t many of such summaries out there, and as a cheat sheet if you want to implement them from scratch.
Developer to Data Scientist
As a developer you would be keenly looking to re-use your development skills into data science and that’s a good thing as development skills are important part of the Data Scientist’s toolbox. …
Like this:
Like Loading…
Related