Introduction to Python Metaclasses
In this tutorial, learn what metaclasses are, how to implement them in Python, and how to create custom ones.
PracticalAI
Empowering you to use machine learning to get valuable insights from data.• Implement basic ML algorithms and deep neural networks with PyTorch.• Run everything on the browser without any set up using Google Colab.• Learn object-oriented ML to code for products, not just tutorials.
Exotic link functions for GLMs
In my previous post on GLMs, I discussed power link functions. But there are much more links that can be used
How different are conventional programming and machine learning? Explained with a toy example
Engineering allowed us to push the limits of human capabilities. We used our understanding of nature and utilized that to serve our purposes. Be it a high performant mechanical machinery or an encoded silicon chip. Computers have been by far one of the most intricate utilization of nature’s forces put to help humans in pushing their limits of capabilities i.e. many tasks which can be performed by computers can never be performed that quickly and efficiently by a human or a set of humans. As Steve Jobs would say, computers are like a bicycle for our minds.
Confidence intervals for GLMs
You’ve estimated a GLM or a related model (GLMM, GAM, etc.) for your latest paper and, like a good researcher, you want to visualise the model and show the uncertainty in it. In general this is done using confidence intervals with typically 95% converage. If you remember a little bit of theory from your stats classes, you may recall that such an interval can be produced by adding to and subtracting from the fitted values 2 times their standard error. Unfortunately this only really works like this for a linear model. If I had a dollar (even a Canadian one) for every time I’ve seen someone present graphs of estimated abundance of some species where the confidence interval includes negative abundances, I’d be rich! Here, following the rule of ‘if I’m asked more than once I should write a blog post about it!’ I’m going to show a simple way to correctly compute a confidence interval for a GLM or a related model.
10 R functions for Linux commands and vice-versa
This post will go through 10 different Linux commands and their R alternatives. If you’re interested in learning more R functions for working with files like some of those below, also check out this post.
How to tune a BigQuery ML classification model to achieve a desired precision or recall
BigQuery provides an incredibly convenient way to train machine learning models on large, structured datasets. In an earlier article, I showed you how to train a classification model to predict flight delays.
Implementing Defensive Design in AI Deployments
With the upcoming launch of one of our AI products, there has been a repeating question that clients kept asking. This same question also shows up once in a while with our consulting engagements, to a lesser degree, but still demands an answer. The simple version of the question is this: How can I know that the AI is doing a good job? Now it’s easy to throw around confusion matrices and neural activation graphs to clients, but they have a much deeper question?-?and a very valid concern. They are not asking about the performance of the system, they are asking about its alignment to their own problems. If this model is now in charge of one or many of their business processes, how can they manage it if they cannot see its criteria for how it is executing its tasks? This touches on a combination of management fundamentals, business logic, and the ongoing evolution of the machine learning fields. The goal of bespoke AI solutions is to accelerate key processes that can alleviate the workload of the rest of the staff, or to make decisions in real time. As such, a system that cannot reliably execute a process within a trustworthy tolerance range might as well not be implemented at all.
6 Emotionally Rewarding Data Science Projects
The field of data science is best-suited for those who love mathematics and working with numbers. While some projects are tedious and monotonous, particularly on the entry level, there are plenty of exciting and rewarding jobs in the sector for qualified, experienced professionals. The dawn of big data and next-gen data analytics makes the field even more innovative and exciting by giving individuals access to more data than ever before. Since we are currently living in the Information Age, it only makes sense to use this data in fun, creative and rewarding ways.
A ‘short’ introduction to model selection
In this post I will discuss a topic central to the process of building good (supervised) machine learning models: model selection. This is not to say that model selection is the centerpiece of the data science workflow?-?without high-quality data, model building is vanity. Nevertheless, model selection plays a crucial role in building good machine learning models.
A review of recent reinforcment learning applications to healthcare
The application of machine learning to healthcare has yielded many great results. However, the vast margin of these focus on diagnosing or forecasting, and not explicitly on treatment. Although these can indirectly help at treating people (for example diagnosis is the first step to finding treatment) in many cases particularly where there are many available treatment options figuring out the best treatment policy to use for a particular patient is challenging for human decision makers. Reinforcement learning has grown quite popular, however the majority of papers focus on applying it board or video games. RL performed well at learning the optimal policies in these(video/board games) contexts but has been relatively untested in real world environments like healthcare. Naturally, RL is a good candidate for this purpose, however there are many barriers for it to work in practicallity. In this article I’m going to outline some of the more recent approaches as well as some of the barriers that still exist with the application of RL to healthcare. If this topic interests you I will also go into more detail about some of these models at the PyData Orono Meetup on Reinforcement Learning in the Real World which will be broadcast on Zoom this Wednesday 7-9:30 EST. This article assumes that you have a basic knowledge of reinforcement learning. If you don’t, I suggest reading one of the many articles already on Towards Data Science on the subject.
Chatbots are cool! A framework using Python Part 1:Overview
The bot framework is modularized which opens up an array of opportunities for the readers to design and implement their own features. Integrations can be done easily in the framework. Also, the probability for failure is minimal since it is designed to be plug and play.Beginner: An overall idea on how the framework is developed and used for this specific project. You should be able to download the codes from Github and complete the setup successfully. This includes package installations, slack and IBM Watson account creation and setup, run one time files to generate the links and movie recommendations. You can add extra skills in IBM Watson (like a small talk which generate static responses) and see the results in slack environment.Intermediate: You should be able to use this framework as a template to design your own chatbot which can be deployed on a different domain. In addition, you can extend the knowledge base for the chatbot by adding new data sources which includes writing codes to connect to different databases (Elastic search, SQL databases, Excel and so on..). Also, you can add extra NLP features to the bot and see the results in slack environment.Expert: You should be able to add/extend bot features by integrating API connections for Slack/NLP. I used IBM Watson to identify question category and to generate static responses. You can replace IBM Watson in the framework by designing your own NLP capabilities. Also, you can extend the bot integrations for different platforms (Web, Skype and so on..)
How should we define AI?
As you have probably noticed, AI is currently a ‘hot topic’: media coverage and public discussion about AI is almost impossible to avoid. However, you may also have noticed that AI means different things to different people. For some, AI is about artificial life-forms that can surpass human intelligence, and for others, almost any data processing technology can be called AI. To set the scene, so to speak, we’ll discuss what AI is, how it can be defined, and what other fields or technologies are closely related. Before we do so, however, we’ll highlight three applications of AI that illustrate different aspects of AI.
The Hidden Dangers in Algorithmic Decision Making
The quiet revolution of artificial intelligence looks nothing like the way movies predicted; AI seeps into our lives not by overtaking our lives as sentient robots, but instead, steadily creeping into areas of decision-making that were previously exclusive to humans. Because it is so hard to spot, you might not have even noticed how much of your life is influenced by algorithms. Picture this?-?this morning, you woke up, reached for your phone, and checked Facebook or Instagram, in which you consumed media from a content feed created by an algorithm. Then you checked your email; only the messages that matter, of course. Everything negligible was automatically dumped into your spam or promotions folder. You may have listened to a new playlist on Spotify that was suggested to you based on the music that you’d previously shown interest in. You then proceeded with your morning routine before getting in your car and using Google Maps to see how long your commute would take today. In the span of half an hour, the content you consumed, the music you listened to, and your ride to work relied on brain power other than your own?-?it relied on predictive modelling from algorithms. Machine learning is here. Artificial intelligence is here. We are right in the midst of the information revolution and while it’s an incredible time and place to be in, one must be wary of the implications that come along with it. Having a machine tell you how long your commute will be, what music you should listen to, and what content you would likely engage with are all relatively harmless examples. But while you’re scrolling through your Facebook newsfeed, an algorithm somewhere is determining someone’s medical diagnoses, their parole eligibility, or their career prospects. At face value, machine learning algorithms look like a promising solution for mitigating the wicked problem that is human bias, and all the ways it can negatively impact the lives of millions of people. The idea is that the algorithms in AI are capable of being more fair and efficient than humans ever could be. Companies, governments, organizations, and individuals worldwide are handing off decision-making for many reasons?-?it’s more reliable, it becomes easier, it is less costly, and it’s time-efficient. However, there are still some concerns to be aware of.
Who Do We Blame When an AI Finally Kills Somebody
We’re rapidly approaching the point where AI will be so pervasive that it’s inevitable that someone will be injured or killed. If you thought this was covered by simple product defect warranties it’s not at all that clear. Here’s what we need to start thinking about.
Like this:
Like Loading…
Related