The Personal MS(DS) is an initiative to customize the Data Science Masters roadmap according to one's interests hence providing complete autonomy to the learner. The intuition behind #thepersonalmsds is to upgrade skills without formally enrolling into a Master's program at a University - EshbanTheLearner/thepersonalMSDS-v2

We experience touch in our daily lives, and are witness to the power of it during poignant moments. We saw this when the principal of Stoneman Douglas High School promised a hug to every student after a school shooting in 2018, when former President Obama embraced the families of the kids at Sandy Hook in 2012, and when Princess Diana broke royal protocol to hug an HIV-positive child in 1991. Time and time again, we see touch wielding incredible power. It remains integral to our communication, even in an age of integrated technology – and it imparts emotion when verbal communication isn't enough. During the 1960s, thousands of Romanian children were thrown into orphanages where they grew up starved of human contact.

Using the NASA moon rover as their inspiration, robotics students at Lake City Secondary School are nearing completion of a working, remote-controlled, wheeled robot rover of their own design. The rover, named The Helios Vulturem (Helios was the Greek titan of the sun, and Vulturem is Latin for Falcon -- the school's mascot and namesake) was designed by Grade 12 robotics students Nathan Cisecki, Jonathan Wolfe, Jayden Guichon, Eric DeVuyst and Colby Ostrom under the guidance of robotics teacher Nick MacDonald, along with some assistance from Grade 11/12 metalwork students Cole Rochefort and Cameron Smithson. The students began the year disassembling a broken-down, miniature quad to eliminate broken parts and repair salvageable ones, while the metalwork students welded the broken steering column and assisted in the reconfiguration of the chassis. "We'd seen this quad sitting up there on a shelf in Mr. MacDonald's class in Grade 10, and he mentioned to us it would be our senior year project, but we never thought we'd be able to make it into a rover," Ostrom said. Four years has led up to this," Cisecki added.

Craig Smith is a former New York Occasions correspondent and host of the podcast, Eye on AI. The continuing Covid-19 pandemic, which has disrupted classroom instruction all over the world, presents an historic alternative to democratize schooling with artificial intelligence. Since Socrates taught Plato and Plato taught Aristotle -- or Confucius taught Yan Hui -- man has identified that one of the best schooling is delivered one-to-one by an skilled educator. The result's the imperfect classroom-based instruction that we stay with in the present day. But, new types of AI, based mostly on deep neural networks, can now uncover patterns about how college students carry out and assist lecturers optimize their methods accordingly.

With the increasing requirement for more data scientists, ML experts, and AI engineers in every industry, Microsoft, in partnership with Netflix, has launched three new learning modules to guide learners through beginning concepts in data science, machine learning and artificial intelligence. Inspired by the new Netflix original film -- 'Over the Moon' these learning modules include three paths -- planning a Moon mission using the Python Pandas Library; predicting meteor showers using Python and VC Code; and using AI to recognise objects in images using Azure Custom Vision. The growing requirement of data scientists comes with criteria of having a broad set of skills from data analysis with no-code and low-code solutions which will help them with designing and writing intricate ML models and solve some of the planet's most difficult problems. Keeping this in mind, Microsoft, partnering with Netflix, has launched these modules for providing high quality, free content to help learners develop their skills depending based on their professional goals and personal interests. According to Microsoft's blog post, "One such endeavour in creating an opportunity for you to learn and upskill is through unique partnerships. In the summer of 2020, we launched a set of Microsoft Learn modules inspired by real NASA scientists and engineers at https://aka.ms/LearnInSpace. And this Fall we are excited to bring you three more Microsoft Learn modules inspired by the new Netflix Original Over the Moon."

Artificial intelligence has taken the world around us by storm, and like all things in the 21st century, it has thoroughly impacted the education sector. From chatbots to automated thermostat conditioning, the machines around us learn from our behavior and then act intelligently to deliver the best possible service or product. Almost every industry out there is leveraging the power of AI. How, then, can education not be impacted by something so all-consuming. AI has changed the delivery, approach, and stylization of the education sector around the globe with relevant innovation.

You don't have to cook or spend any of your hard-earned money. The truth is unless if you count special talks and lectures in graduate school that promise free pizza, there is no free lunch in machine learning. The "no free lunch" (NFL) theorem for supervised machine learning is a theorem that essentially implies that no single machine learning algorithm is universally the best-performing algorithm for all problems. This is a concept that I explored in my previous article about the limitations of XGBoost, an algorithm that has gained immense popularity over the last five years due to its performance in academic studies and machine learning competitions. The goal of this article is to take this often misunderstood theorem and explain it so that you can appreciate the theory behind this theorem and understand the practical implications that it has on your work as a machine learning practitioner or data scientist.

With data durability, high access speed, low power efficiency and byte addressability, NVMe and SSD, which are acknowledged representatives of emerging storage technologies, have been applied broadly in many areas. However, one key issue with high-performance adoption of these technologies is how to properly define intelligent cache layers such that the performance gap between emerging technologies and main memory can be well bridged. To this end, we propose Phoebe, a reuse-aware reinforcement learning framework for the optimal online caching that is applicable for a wide range of emerging storage models. By continuous interacting with the cache environment and the data stream, Phoebe is capable to extract critical temporal data dependency and relative positional information from a single trace, becoming ever smarter over time. To reduce training overhead during online learning, we utilize periodical training to amortize costs. Phoebe is evaluated on a set of Microsoft cloud storage workloads. Experiment results show that Phoebe is able to close the gap of cache miss rate from LRU and a state-of-the-art online learning based cache policy to the Belady's optimal policy by 70.3% and 52.6%, respectively.

We present DeepMind Lab2D, a scalable environment simulator for artificial intelligence research that facilitates researcher-led experimentation with environment design. DeepMind Lab2D was built with the specific needs of multi-agent deep reinforcement learning researchers in mind, but it may also be useful beyond that particular subfield.

Deep learning classifiers are assisting humans in making decisions and hence the user's trust in these models is of paramount importance. Trust is often a function of constant behavior. From an AI model perspective it means given the same input the user would expect the same output, especially for correct outputs, or in other words consistently correct outputs. This paper studies a model behavior in the context of periodic retraining of deployed models where the outputs from successive generations of the models might not agree on the correct labels assigned to the same input. We formally define consistency and correct-consistency of a learning model. We prove that consistency and correct-consistency of an ensemble learner is not less than the average consistency and correct-consistency of individual learners and correct-consistency can be improved with a probability by combining learners with accuracy not less than the average accuracy of ensemble component learners. To validate the theory using three datasets and two state-of-the-art deep learning classifiers we also propose an efficient dynamic snapshot ensemble method and demonstrate its value.