Menu
Menu
Maya Bercovitch is a Director of Engineering at Anaplan, and is responsible for leading the Data Science department. Anaplan is a global company and a leader in Connected Planning – a unique technology that connects all of the data and plans for the largest companies in the world. The Data Science department is responsible for the AI algorithmic core of Anaplan’s platform, developing AI-based engines and solutions available to Anaplan’s non-data scientist customers. Before joining Anaplan, Maya worked as a data scientist and a DS team leader in a few startups. Overall, she has 12 years of experience in the field of ML and DS. Maya holds an MSc in Information System Engineering from Ben-Gurion University.
Maya Bercovitch is a Director of Engineering at Anaplan, and is responsible for leading the Data Science department. Anaplan is a global company and a leader in Connected Planning – a unique technology that connects all of the data and plans for the largest companies in the world. The Data Science department is responsible for the AI algorithmic core of Anaplan’s platform, developing AI-based engines and solutions available to Anaplan’s non-data scientist customers. Before joining Anaplan, Maya worked as a data scientist and a DS team leader in a few startups. Overall, she has 12 years of experience in the field of ML and DS. Maya holds an MSc in Information System Engineering from Ben-Gurion University.
Whether it is the monthly demand of a specific commodity over the next year, the weekly financial performance of a company in the upcoming quarter, or the daily spread of COVID-19 during the next 2 months – time series algorithms can be used to predict the future. Just like in other fields of Data Science, defining well appropriate evaluation metrics is the most important matter a data scientist needs to tackle in order to enhance their prediction accuracy. A suitable metric should represent the customer’s objectives – as different objectives can lead to choosing a different algorithm or configuration.
In this lecture, we will discuss the basic differences between some of the commonly used metrics in the field of Time Series Forecasting and we will see how the use of different metrics affects the overall evaluation of our model.
Whether it is the monthly demand of a specific commodity over the next year, the weekly financial performance of a company in the upcoming quarter, or the daily spread of COVID-19 during the next 2 months – time series algorithms can be used to predict the future. Just like in other fields of Data Science, defining well appropriate evaluation metrics is the most important matter a data scientist needs to tackle in order to enhance their prediction accuracy. A suitable metric should represent the customer’s objectives – as different objectives can lead to choosing a different algorithm or configuration.
In this lecture, we will discuss the basic differences between some of the commonly used metrics in the field of Time Series Forecasting and we will see how the use of different metrics affects the overall evaluation of our model.
© 2018-2021 WiDS TLV – Intuit. All rights reserved.
Scotty – By Nir Azoulay
Design: Sharon Geva
