With students from a range of science, technology, engineering, and math (STEM) competitions from across the country looking on, U.S. President Barack Obama delivers remarks after viewing science projects at the White House Science Fair, at the White House, March 23, 2015 in Washington, DC. (Photo: Drew Angerer/Getty Images) It is already a given fact that gender inequality still continue to persist in the field of education. Despite the government's efforts to ensure that all students should have access to high quality education, gender gap remain notable, particularly in STEM (Science, Technology, Engineering And Mathematics)-related and CTE (Career and Technical Education) curricula. Fortunately, artificial intelligence (AI) has been considered as a powerful tool in bridging the gender gap in STEM education. That's why, Stanford has launched a tuition-free AI camp called SAILORS to encourage young girls, as well as "underrepresented minorities" to explore STEM-related fields. Initially launched on the summer of 2015, Stanford Artificial Intelligence Outreach Summer aka SAILORS was created by computer science professor Fei-Fei Li and Postdoc (postdoctoral scholar) Olga Russakovsky.

I recently participated in a weekend-long data science hackathon, titled'The Smart Recruits'. Organized by the amazing folks at Analytics Vidhya, it saw some serious competition. Although my performance can be classified as decent at best (47 out of 379 participants), it was among the more satisfying ones I have participated in on both AV (profile) and Kaggle (profile) over the last few months. Thus, I decided it might be worthwhile to try and share some insights as a data science autodidact. The competition required us to use historical data to create a model to help an organization pick out better recruits. The evaluation metric to be used for judging the predictions was AUC (area under the ROC curve).

Ozobot's Bit impressed us a few years ago with its simply take on programming education: kids just need to draw lines on a piece of paper or mobile device to program the tiny robot. As they get more comfortable, they can start to program on mobile devices and computers. Now Ozobot is taking a major step forward with the 100 Evo, a new robot that has sensors to interact with its environment, lights, a speaker and social capabilities. While Ozobot's previous devices were aimed directly at kids, it's hoping that Evo can break through to high schoolers and even college students, according to founder and CEO Nader Hamda. The new bot has a shot of appealing to older students simply because it can do a lot more than before.

During the christmas vacation of 2015, I got a motivational boost again and decided try out Kaggle. So I spent quite some time experimenting with various algorithms for their Homesite Quote Conversion, Otto Group Product Classification and Bike Sharing Demand contests. The main takeaway from this was the experience of iteratively improving the results by experimenting with the algorithms and the data. I learned to trust my logic when doing machine learning. If tweaking a parameter or engineering a new feature seems like a good idea logically, it's quite likely that it actually will help.

We can see a few interesting things in this table. In a similar talk by Owen at ODSC Boston 2015 titled "Open Source Tools and Data Science Competitions", he again summarized common parameters he uses: We can see some minor differences that may be relevant. Finally, Abhishek Thakur, in his post titled "Approaching (Almost) Any Machine Learning Problem" provided a similar table listing out key XGBoost parameters and suggestions for tuning. The spreads do cover the general defaults suggested above and more. It is interesting to note that Abhishek does provides some suggestions for tuning the alpha and beta model penalization terms as well as row sampling. You can develop and evaluate XGBoost models in just a few lines of Python code.

In an era of ubiquitous large-scale streaming data, the availability of data far exceeds the capacity of expert human analysts. In many settings, such data is either discarded or stored unprocessed in datacenters. This paper proposes a method of online data thinning, in which large-scale streaming datasets are winnowed to preserve unique, anomalous, or salient elements for timely expert analysis. At the heart of this proposed approach is an online anomaly detection method based on dynamic, low-rank Gaussian mixture models. Specifically, the high-dimensional covariances matrices associated with the Gaussian components are associated with low-rank models. According to this model, most observations lie near a union of subspaces. The low-rank modeling mitigates the curse of dimensionality associated with anomaly detection for high-dimensional data, and recent advances in subspace clustering and subspace tracking allow the proposed method to adapt to dynamic environments. Furthermore, the proposed method allows subsampling, is robust to missing data, and uses a mini-batch online optimization approach. The resulting algorithms are scalable, efficient, and are capable of operating in real time. Experiments on wide-area motion imagery and e-mail databases illustrate the efficacy of the proposed approach.

As Torrance observes: '[c]reativity defies precise definition... even if we had a precise conception of creativity, I am certain we would have difficulty putting it into words' [15, p. 43]. Many other authors have expressed similar difficulties [7, 10, 16]. In their review of research into human creativity, Hennessey and Amabile ask a significant follow-on question: 'Even if this mysterious phenomenon can be isolated, quantified, and dissected, why bother? Wouldn't it make more sense to revel in the mystery and wonder of it all?' [11, p. 570] Two answers to this question are offered by Hennessey and Amabile, both of which are identified as desirable: to gain a deeper understanding of creativity and to learn how to boost people's creativity. Creativity can and should be studied and measured scientifically, but the lack of a commonly-agreed understanding causes problems for measurement [10]. Plucker et al. make recommendations about best practice based on their own survey of the creativity literature: 'we argue that creativity researchers must (a) explicitly define what they mean by creativity, (b) avoid using scores of creativity measures as the sole definition of creativity (e.g., creativity is what creativity tests measure and creativity tests measure creativity, therefore we will use a score on a creativity test as our outcome variable), (c) discuss how the definition they are using is similar to or different from other definitions, and (d) address the question of creativity for whom and in what context.' [9, p.92] In short, we need to specify and justify the standards that we use to judge creativity. A more objective and well-articulated account of how creativity is manifested enables researchers to make a worthwhile contribution [8-10]. Particularly, in research we would like to focus on what processes and concepts relevant to creativity are'sufficiently important to warrant study' [17, p. 15], based on an accumulation of the body of work on creativity to date [17].

Nearly 50 million students are now returning to classrooms--from kindergarten through 12th grade. They will spend approximately eight hours a day at school and additional hours doing homework. They will be educated, in public schools alone, by the equivalent of over 3 million full-time teachers. And they will, with rare exception, learn a dismal fraction of what they ought to be learning to be creative, confident and critical thinkers about themselves and the world around them. As a parent myself, I literally apologized to each of my children--and not just once--for the fact that so much of their time as grade school and junior high school and high school students (even at private school) was being spent on memorization, regurgitation and rote learning that amounted to busy work and the warehousing of them, physically and mentally.

Four UC Berkeley researchers developed a program to help grade papers during their time working as teaching assistants – and now, they've added artificial intelligence to their app to help instructors speed up the grading process. The team launched the online grading app Gradescope two years ago and have accumulated 10 million answers to around 100,000 questions from a wide range of college courses – the app has already shortened the grading process by 50 percent due to its friendly interface and the ability for multiple teaching assistants to grade papers in parallel. Their new AI features addresses three challenges: identify question types, distinguishing between different written marks, and recognizing handwriting. AI helps turn grading into an automated, highly repeatable exercise by learning to identify and group answers, and thus treat them as batches. The addition of AI promises to slash grading times by as much as 90 percent, said Sergey Karayev, a Gradescope co-founder who finished his PhD in computer science in 2014.

During the christmas vacation of 2015, I got a motivational boost again and decided try out Kaggle. So I spent quite some time experimenting with various algorithms for their Homesite Quote Conversion, Otto Group Product Classification and Bike Sharing Demand contests. The main takeaway from this was the experience of iteratively improving the results by experimenting with the algorithms and the data. I learned to trust my logic when doing machine learning. If tweaking a parameter or engineering a new feature seems like a good idea logically, it's quite likely that it actually will help.