Webinar du 22.04.20 – Abakus SIGN & IBM i2 Analyst Notebook investigation graphique

Webinar en partenariat entre Satom IT & Learning Solutions et Swissintell.

Présentation du connecteur Abakus SIGN sur une base Graph des Panama Papers pour présentation de relations cachées des résultats Investigation graphique à l’AI.

Recommendations for an AI Strategy in Switzerland

Digital transformation is radically reshaping almost every aspect of our society. The explosion of artificial intelligence (AI) and big data analytics applications is enabled by the extreme availability of data in combination with the substantial computing power of modern highly distributed computing infrastructures connected by high-speed networks. Machine learning technologies can be trained to perform specific tasks with an efficiency and an accuracy that can supplement and, in some cases, outperform that of humans. These systems provide deep insights by learning from data and interactions with users, which is already leading to a profound transformation of numerous industries, professions and society at large. The current state of AI is, however, still far from delivering truly intelligent behaviour that is comparable to human intelligence. An AI research strategy should therefore carefully analyse AI’s history with its various waves of large promises and conceptual shortcomings.

Recent advancements in machine learning have enabled AI technologies to become extremely successful. Speech recognition, natural language interaction with machines and facial recognition based on deep learning are now commodities that have changed the way people interact. The machine learning strategy of emulating human performance by learning from human experience promises a solution to the knowledge extraction problem. However, the automated reasoning process is as opaque as human decision making. Evolution has enabled humans to collectively reason and act on our collective experience, though other humans are often black boxes. Today, we are confronted with computational artefacts that are adapted to complex human decision making and, thereby, have inherited a similar “black box” behaviour.

Given the penetration of AI across most industries, its potential impact on GDP promises to be very high. In Switzerland, AI is already reshaping industries such as banking, insurance, pharmaceuticals and manufacturing. Furthermore, Switzerland is the European country that has the highest number of AI start-ups per citizen, with more than 100 startups. Many leading countries are heavily investing in AI development strategies and the establishment of technology transfer centres in this field.

To date, Switzerland has not developed a dedicated AI strategy. AI is one of many topics covered in the strategy “Digitale Schweiz”. An interdepartmental working group on AI which should ensure knowledge exchange in the domain of AI within the federal administration and coordinate Switzerland’s positions in international bodies, is mandated to submit a report to the Federal Council by September 2019. Furthermore, an interdisciplinary study on behalf of TASWISS is evaluating the opportunities and risks of AI on the basis of various focal points: work, education, media, consumption and administration. The publication of that study is planned for the end of 2019.

Read more here : SATW

Quantifying the Accuracy of Mobility Insights from Cellular Network Data

From mobile-network data to mobility indicators

The mobility insights big-data platform processes anonymised network events — more than 2 Million per second — and produces more than 10 Million anonymous trips per day. These trips are trajectories that we describe in term of time, space and mode of transport. As we focus on collective mobility, we aggregate these trips into dynamic mobility indicators that describe minute by minute the mobility pulse of Switzerland. For example, we are able to quantify, for a given minute of the day, the number of highway and train trips that go through a given area. We are also able to build the associated distribution of origins and destinations. All results we share are k-anonymised to minimise the risk of re-identification.

Collecting ground-truth data

Benchmarking machine-learning algorithms requires data that associate samples with ground truth (actual label). This is challenging given that the machine learning task at hand is very specific with no public datasets available. We therefore decided to collect the data ourselves: we developed an application — only open to Swisscom employees with an explicit opt-in — that provides personalised mobility reports which describe the daily trips performed by the user as well as the associated CO2 footprint. At the same time, the user is able to provide feedback: She can rate the reconstructed trip, correct the origin and destination of each trip as well as the detected mode of transport.

Source : Medium & Swisscom