Data and analytics

Time Series Forecasting with FBProphet

Traditional approaches like SARIMA models often require manual data pre-processing steps (e.g. differencing to make the data stationary). In addition, these models are not allowed to add additional domain knowledge to improve precision. For solving these problems, Facebook researchers recently released FBProphet.

Using data to drive improvements through experimentation

Much of the world’s data is generated online, which has the advantage of being easy to collect. However, before data can be used, it needs to be turned into information.

Using Statsmodels GLMs to model Beverage Consumption

The use of Python for data science and analytics is growing in popularity and one reason for this is the excellent supporting libraries (NumPy, SciPy, pandas, Statsmodels (ref), Scikit-Learn, and Matplotlib, to name the most common ones). One obstacle to adoption can be lack of documentation: e.g. Statsmodels documentation is sparse and assumes a fair level of statistical knowledge to make use of it. This article shows how one feature of Statsmodels, namely Generalized Linear Models (GLM), can be used to build useful models for understanding count data.

What Spark did next: A case study in product maturation

When Apache Spark 1.0 was released in mid-2014, it was quickly recognised for its compelling rethink of the conventions of large-scale data processing. Stepping away from the then prevalent MapReduce mindset, Spark introduced the concept of the Resilient Distributed Dataset (RDD); an immutable data multiset distributed over a cluster.