Dashboards are a convenient method for quickly present data. For data scientists and analysts, it’s the tool of choice for showing results of their work to the non-technical recipients. Of course, there are many services for making online dashboards with no or little code. Using ready-made solutions such as Tableau, Qlikview or Google Data Studio, quick production of decent looking charts and dashboards is not a problem.

However, most of them are too expensive. We also often encounter situations where the possibilities for adapting views to exceptional data are limited. On the other hand, tailored, self-hosted solutions cause other difficulties. …

Conjoint analysis is a market research method used to measure customer preferences and the importance of various attributes of products or services. In this method, products or services (real or hypothetical) are presented to respondents (e.g. potential consumers) as a set of profiles. Each profile is described by attributes and their levels. Those attributes and their levels are explanatory variables. On the basis of the collected assessments (preferences) from consumers, a breakdown (decomposition) of total preferences is made, using statistical methods, by calculating the share of each attribute in the estimated total utility of the profile.

This way, marketers can identify which combination of attributes is going to perform on the market or which will be the most sought after combination in terms of the consumer’s point of view before actually introducing a new product. …

What if a website would be able to predict your preferences and adjust its appearance to maximize your experience?

Or if a web-based video conferencing application could replace a messy room with another background and cancel noises of your dog?

We already have the technology.

Thanks to JavaScript machine learning frameworks, such as Tensorflow.js or ONNX.js, we can easily implement these functions in the web environment.

However, they have one disadvantage — poor performance in comparison with native apps.

Neural hardware

Hardware architectures specialized for machine learning purposes emerged because of increasing demand for computing capabilities.

Meteogram is a graphical presentation of meteorological variables (observed or forecasted) through time, for a specific location. The meteogram below was created by combining several forecasts from Meteoblue’s newsletter.

Each “frame” contains 5-day meteogram with 3 simple graphs for a specific location:

• Temperature chart with weather pictograms. The time from sunrise to sunset is indicated in light yellow.
• Clouds in different altitudes: from a few clouds (light grey) to overcast (dark grey). Dark blue bars show hourly precipitation and light blue showers. An asterisk indicates snowfall.
• Forecasts for wind speeds are blue and for gusts are green. …

If I had to choose one basic skill in data science that is the most useful, it would be time series forecasting. Predicting the future value of something contributes to making better decisions. Therefore, it is crucial to be sure that you can rely on forecasting. The choosing, construction and interpretation of forecast evaluation statistics are just as important as making forecasts.

What actually makes up the accuracy of the forecast?

How to choose evaluation statistics

Choosing a performance estimation method often depends on the domain of the problem. In my career, I have encountered a situation where a hastily chosen metric has caused the client’s dissatisfaction with the forecasting results optimized for KPI inconsistent with the specific business case. …

Surveys are a convenient method for quick collecting people’s opinions in many areas. Despite a giant increase in access to data collected passively (e.g. browsing history, eye tracking), organisations still rely on gathering feedback from their customers. Survey data can be used for sentiment analysis, designing new products, building better recommendation engines and market segmentation. The survey can be conducted in person, by phone or online.

Of course, there are many paid services for making online questionnaires with no or little code. For open-source solutions, the industry standard is LimeSurvey. Getting started with a simple questionnaire is rather quick, but you still need hosting for LimeSurvey. …

In the previous article, I introduced a conjoint analysis and provided some examples of how useful the market research method is. I recommend you to read it first.

Choice-based conjoint analysis (CBC, or: discrete choice modelling, discrete choice experiment, experimental choice analysis, quantal choice models) uses discrete choice models to collect consumer preferences. The main difference distinguishing choice-based conjoint analysis from the traditional full-profile approach is that the respondent expresses preferences by choosing a profile from a set of profiles, rather than by just rating or ranking them. …

Interesting data science project in a portfolio could not only validate your skills but also help find a job. Beginner data scientists often struggle to come up with ideas for their next project.

But there is a way to find ideas and to determine that they are worth time.

Dr Hal R. Varian is the Chief Economist at Google and also an emeritus professor at the University of California, Berkeley. Maybe some of you, who studied economics, recognize him as the author of economics textbooks.

In the mid-nineties, Varian described his method of making economic models in the popular essay. …