WiDS Stockholm 2021

WiDS Stockholm is an independent event that is organized by WiDS Sweden as part of the annual WiDS Worldwide conference organized by Stanford University and an estimated 150+ locations worldwide, which features outstanding women doing outstanding work in the field of data science. 

This talk will be focused on the importance of setting KPIs (Key Performance Indicators), choosing the right KPIs as well as the power they bring to steer big operational organizations. KPIs are usually a great way to measure the success of features, new initiatives, and overall performance. However, if not defined properly, they won’t tell the true story and can sometimes even trigger unwanted behavior. We will discuss how to define KPIs and what makes them different from other metrics and measurements. This talk will focus on the importance of seeking qualitative insights, to really be able to understand the connection between the KPI and the impact they have on behavior and actual performance.

In today's world of entertainment, the problem is not the lack of content but rather the overwhelming flood of choices leaving the customer in distress. Apart from the sheer number of options, it is also about time and attention: How do you catch your audience in 45 seconds? How do you keep them coming back? Personalized recommendations are one key element in tackling these issues. In this talk, we first take a step back and look into what a recommender system is, and what different types and algorithms there are available today. In this session, we will examine the case of Viaplay (NENT Group's streaming service) and share learnings from our experience from building personalization from the ground up.

COVID Symptom Study in Sweden collects data from some 200 000 volunteers to map and investigate the spread of COVID19. In this talk, we introduce the biostatistics and data science aspects involved in analysing data from the study, among them challenges associated with voluntary data collection, statistical methods used to address biases in our data, and methods to estimate the prevalence of symptomatic COVID. We will also describe data science approaches instrumental to giving us a better understanding of the pandemic and its evolution. Two such examples are how we identify new symptoms and symptom groups from comments and figure out what role the geographical location of a COVID test center might play in testing behaviour, especially in disadvantaged communities.

AI in its general term is rapidly evolving internationally and becoming a global priority. Harnessing AI opportunities highly relies on efficient hardware platforms that could enable the wide deployment of applications, especially those relying on deep neural networks (DNN). DNNs are opening their routes in different domains such as e-health, robotics, and autonomous driving. For many of these applications, the processing must happen fast, locally, and with low power consumption, which demands powerful computing platforms. In this presentation, we will talk about different hardware platforms to execute DNNs, learn the science/engineering gap in this area, and investigate methods and approaches to enable dreams (DNN-enabled applications) to become a reality (in our everyday devices)!

The AI ecosystems have been growing and changing continuously in recent years. This talk aims to describe the key participants within the Swedish ecosystem, from industry and startups to public sector, academia and funding players. The aim of this talk is to draw a map of the state of AI in Sweden: What is the role of WiDS Sweden in it? What is YOUR role in it and what paths and opportunities might be available and interesting to you? We also touch on the interesting trends and important questions in the AI landscape.

6 talks, each, 5 minutes of presentation and 5 minutes of questions.

• Celine Xu, Lead Data Scientist at Axel Johansson: The challenge of unbalanced data in recommendations systems
• Helga Westerlind, Assistant Professor at Karolinska University: Harmonization of data across countries and study types
• Barbara Livieri, Senior Data Scientist at Spotify: When life gives you imprecise exposure, you make causal inference
• María García, Data Scientist at IKEA: Affordability index in price recommendation models
• Kunru Chen, PhD student in Machine Learning at Halmstad University: Semi-supervised Representation Learning for Machine Activity Recognition
• CJ Jenkins, Senior Data Scientist at Klarna: Encoding tricky data: high cardinality in categorical variables and text variables make for complicated feature encoding

In our career panel, we will discuss work in the industry and academia, the broad collection of roles which belong under the data science umbrella, career paths to and through data science, whether individual contributor or manager roles might be for you, and what your workplace can do for you (and not just what you can do for your workplace). And, as usual, take many many of questions!

Panelists:

• Luminita Färnström, Tech Lead for the Data Analysis and Insights team at SL
• Marlena Maziarz, Biostatistician at Lund University Diabetes Centre
• Maddy Renström, Data Science Team Lead at King

Moderate by Galina Esther Shubina

Join us for interactive fun at gather.town, where you can chat with the attendees, speakers, organizers and sponsors. As with our last WiDS conference, the mingle is "do not miss" fun.