Data Analytics

Counting on a multidisciplinary team, with different backgrounds and professional experiences, BioDecision Analytics offers data analysis services for both regulated and unregulated studies, combining statistical methods and modeling techniques. For this, we use Data Analytics and Business Intelligence (BI) tools, aiming to integrate and transform in vitro, preclinical (in vivo), and clinical data into useful information for the development of biotechnological and (bio)pharmaceutical products in general. Through parametric and non-parametric statistical analyses with the most modern machine learning techniques, we can explore the raw data, extracting information that allow us to better understand the investigational product (IP) in development and facilitating the registration process.

How we do?

Analytics represent the extensive investigation of raw data and/or statistics to gain insight and derive value from the data guiding better decision making. It involves the combined use of several quantitative tools to process and evaluate data, identify patterns and behaviors, and subsequently draw conclusions. The process is divided into sequential steps that aim, firstly, at the acquisition and selection of the different sources of most relevant data for the analyses. Subsequently, the raw data are processed, grouped, and treated to obtain an adequate data base for statistical. The workflow Analytics é mostrado abaixo:

Analytics Image

There are several forms of Data Analytics, including:

Descriptive analytics: Comprises a set of techniques and tools that allow summarizing, describing, and characterizing data. Furthermore, it involves calculating your statistics and creating visualization forms suitable for reports that empower decision makers. In a business data context, it can often be defined as Business Intelligence.
Predictive analytics: Uses a set of statistical techniques or machine learning algorithms in order to analyze historical data and, through them, make predictions about data of interest or future behavior. Machine learning is a segment of the Artificial Intelligence area that comprises the use of algorithms that progressively learn from data, thus becoming more accurate in identifying relationships and making predictions of a phenomenon of interest. The techniques involved depend on the type of data available and the objectives of the analysis and can be divided into two broad approaches: Supervised and unsupervised. In the supervised approach, the data have a target variable which os is interested in predicting and analyzing. The unsupervised approach, however, does not involve a target of interest and the algorithm looks for patterns within the data set. In the supervised approach, there is a wide range of techniques depending on the type of study.

A general map of the main algorithms is given below:

ML Image