We often hear in the media that research has proved that “insert any hypothesis”. We would like to show what happens with data behind those studies, also to invoke critical thinking. Statistical significance, p_value and other terms needed to draw conclusions from data will be explained in simple language. Comparison will be drawn between machine learning techniques that are about predictions and statistical procedures when reasons and conclusions are searched for.
What will you learn
- Ways to approach proving or disproving your hypothesis in a formal way
- Pitfalls to be aware of in hypothesis testing
What will you try
You will try to construct your own hypothesis and to select which data you need to collect to prove it. You will also learn how to think critically about cases presented and answer questions about them.
What should you already know and be able to do
Basic statistics and mathematics knowledge would be helpful. Open and curious mind is definitely a must. No software installation or registration is a prerequisite for the Tech Meetup. The Tech Meetup will be held in English.
For people who are interested in data science. It provides an introduction/ways of thinking about certain data science topics for beginners.
For people who already have some experience, the series will be beneficial by hearing simple language explanations about various data science topics, the same language could be reused in the communication with business stakeholders.
How will it be
The MSD Data Scientists will be present hypothesis testing examples from real world scenarios together with pitfalls to demonstrate that sometimes it is not easy to get to any conclusion. Each example will contain not only information but also quiz style questions provoking participants to think about the cases themselves.
- Justina Ivanauskaite: Justina is a senior data scientist with background in statistics. Her expertise lies primarily in econometric modeling, statistics and simulation. She has applied her knowledge in wide range of projects: vaccine research, forecasting, promotion response modeling, or animal health. Justina is interested in creating data science solutions that bring value to the client, with emphasis on reusability and reproducibility. She likes to explore business challenges and suggest methodologies how to achieve desired targets using data science.
- Thomas Browne: Thomas is a Senior Data Scientist originally from Paris, France, where he graduated with a Ph.D. in Statistics. Currently working at Organon, former Data Scientist in MSD, Thomas has worked throughout his young career with mathematical and machine learning models to address complex issues from diverse fields: nuclear energy, pharmaceutical industry as well as theoretical research. He cultivates a real passion for communication and knowledge sharing which implies explaining complicated solutions in the simplest terms.
- Petr Hrobar: Petr is a Data Scientist in MSD with a main background in statistics and econometrics. In his data science work, he used his knowledge to help the researchers on a variety of projects: developing shiny tools, promotion response modeling as well as projects within the animal health division. Petr is interested in multiple fields of data science, with an emphasis on an appropriate methodological framework and strong knowledge of used models.