Microfluidic devices are increasingly used in biological and chemical experiments due to their cost-effectiveness for rheological estimation in fluids. However, these devices often face challenges in terms of accuracy, size, and cost. This study presents a methodology, integrating deep learning, modeling and simulation to enhance the design of microfluidic systems, used to develop an innovative approach for viscosity measurement of polymer melts. We use synthetic data generated from the simulations to train a deep learning model, which then identifies rheological parameters of polymer melts from pressure drop and flow rate measurements in a microfluidic circuit, enabling online estimation of fluid properties. By improving the accuracy and flexibility of microfluidic rheological estimation, our methodology accelerates the design and testing of microfluidic devices, reducing reliance on physical prototypes, and offering significant contributions to the field.

Federated Learning, as a popular paradigm for collaborative training, is vulnerable against privacy attacks. Different privacy levels regarding users' attitudes need to be satisfied locally, while a strict privacy guarantee for the global model is also required centrally. Personalized Local Differential Privacy (PLDP) is suitable for preserving users' varying local privacy, yet only provides a central privacy guarantee equivalent to the worst-case local privacy level. Thus, achieving strong central privacy as well as personalized local privacy with a utility-promising model is a challenging problem. In this work, a general framework (APES) is built up to strengthen model privacy under personalized local privacy by leveraging the privacy amplification effect of the shuffle model. To tighten the privacy bound, we quantify the heterogeneous contributions to the central privacy user by user. The contributions are characterized by the ability of generating "echos" from the perturbation of each user, which is carefully measured by proposed methods Neighbor Divergence and Clip-Laplace Mechanism. Furthermore, we propose a refined framework (S-APES) with the post-sparsification technique to reduce privacy loss in high-dimension scenarios. To the best of our knowledge, the impact of shuffling on personalized local privacy is considered for the first time. We provide a strong privacy amplification effect, and the bound is tighter than the baseline result based on existing methods for uniform local privacy. Experiments demonstrate that our frameworks ensure comparable or higher accuracy for the global model.

Understanding the application of AI to business requires an understanding of context -- strategy, customers, company culture, and so forth. One application worthy of study across organizations is wealth management. A number of banks and investment firms are trying to use AI to improve that management -- either to eliminate human wealth advisers altogether or, much more commonly, to augment their efforts. Our survey research suggests that while many organizations have challenges with production deployments of AI, wealth management is a clear exception. We've studied wealth management strategies using AI and interviewed the analytics and AI officers who support them at several different companies.

Automated source code summarization is a popular software engineering research topic wherein machine translation models are employed to "translate" code snippets into relevant natural language descriptions. Most evaluations of such models are conducted using automatic reference-based metrics. However, given the relatively large semantic gap between programming languages and natural language, we argue that this line of research would benefit from a qualitative investigation into the various error modes of current state-of-the-art models. Therefore, in this work, we perform both a quantitative and qualitative comparison of three recently proposed source code summarization models. In our quantitative evaluation, we compare the models based on the smoothed BLEU-4, METEOR, and ROUGE-L machine translation metrics, and in our qualitative evaluation, we perform a manual open-coding of the most common errors committed by the models when compared to ground truth captions. Our investigation reveals new insights into the relationship between metric-based performance and model prediction errors grounded in an empirically derived error taxonomy that can be used to drive future research efforts

The Dow Jones Industrial Average shed more than 1,300 points earlier this month in the most dramatic drop since February. If you are a client of Morgan Stanley MS Wealth Management, you may have received a message from your financial advisor. The purpose of the message would be to tell you exactly what happened in the market and what the firm's investment professionals are saying about it. It would also tell you your portfolio's current probability of success in light of recent events. The email would come from your financial advisor.

While artificial intelligence-powered tools are being promoted as the next wave of innovation in banking, there is a deep human effort involved in their construction. "You have to have hundreds and hundreds of conversations," with experts from many professional fields, said Jeffrey McMillan, Morgan Stanley's chief analytics and data officer, in a panel at the In Vest Conference. In the traditional way of machine learning, technologists might sift through millions of phone calls to train a bot. Since the bank does not record client calls and the daily customer issues financial advisers face are complex, Morgan Stanley has to find experts who can provide detailed answers to questions from the bank. "You build up a corpus of knowledge," McMillan said.

Bots that suggest investments to your clients in mere seconds.The future may sound a bit spooky for advisors worried about losing their jobs to automatons. The truth is, as complex digital tools become more practical, advisors will soon have to decide when and where to implement them, says American Banker editor-at-large Penny Crosman. Pitfalls and misfires are bound to come with early adoption. How will advisors avoid missteps, all while pioneering new technology? Striking the right balance between humans and technology is becoming the holy grail for advisory firms.

"I believe that by the end of the century, the use of words will have been altered so much that one will be able to speak of machines thinking without expecting to be contradicted." The foundation of AI started back in 1950 when scientist Alan Turing published a seminal paper containing a description of what is now referred to as the "Turing Test" which is designed to determine if a machine can think. A group of scientists got together at Dartmouth College a few years later and coined the term "artificial intelligence". Fast forward to this year and a panel discussion at the Invest 2017 Conference held in New York City on the ability of AI to revolutionize wealth management. Implementing AI technology is an offensive move all the way, Clinc's Mars stated.

Morgan Stanley is about to augment its 16,000 financial advisers with machine-learning algorithms that suggest trades, take over routine tasks and send reminders when your birthday is near. The project, known internally as "next best action," shows how one of the world's biggest brokerages aims to upgrade its workforce while a growing number of firms roll out fully automated platforms called robo-advisers. The thinking is that humans with algorithmic assistants will be a better solution for wealthy families than mere software allocating assets for the masses. At Morgan Stanley, algorithms will send employees multiple-choice recommendations based on things like market changes and events in a client's life, according to Jeff McMillan, chief analytics and data officer for the bank's wealth-management division. Phone, email and website interactions will be cataloged so machine-learning programs can track and improve their suggestions over time to generate more business with customers, he said.

Systems that provide automated investment advice from financial firms have been referred to as robo-advisers. While no one in the industry is particularly fond of the term, it has caught on nonetheless. However, the enhanced human advising process -- augmented by machine learning -- that was recently announced by Morgan Stanley goes well beyond the robo label, and may help to finally kill off the term. New York–based Morgan Stanley, in business since 1935, has been known as one of the more human-centric firms in the retail investing industry. It has 16,000 financial advisors (FAs), who historically have maintained strong relationships with their investor clients through such traditional channels as face-to-face meetings and phone calls.