RivetQuant

Analytical Processes

Our offer in the area of broadly defined analytical processes
includes:

Data Analysis

The most appreciable value that distinguishes Rivet Quant on the market is our experience and diversity, which makes us an interdisciplinary team. Our team members are well-qualified specialists who use following methods in their work.
• • •

01. SUPERVISED MACHINE LEARNING

Regression allows us to characterise and forecast quantitative values ​​of the researched phenomenon. We conduct the regression analysis using both traditional methods (linear regression) and modern methods, such us:



Neural
Networks


Gradient Boosting
Method


Random Forests

Classification allows us to characterise and forecast ​​categorised values.
Also in case of classification methods, our team is able to build models based on the most modern algorithms.
What is more, we can create classification models that allow for interpretation of the classification criteria.

Among others, we use following methods:

Decision
Trees
Rough set-based
methods

02. UNSUPERVISED MACHINE LEARNING CALLED SEGMENTATION

Segmentation allows us to group objects in order to indicate their characteristic feature.
As a company, we conduct this analysis using appropriate statistical methods, such as:

Principal Component
Analysis [ PCA ]
Hierarchical
Clustering

03. Deep learning

In response to challenges which our clients have to meet, such as classification of images or movies, our specialists can build models based on the techniques of deep learning.

An example of using those algorithms could be automatic verification of handwritten postal code.


04. Time series analysis

Time series analysis allows us to predict the value of process based on historical data.
An example of using those methods is the prediction of changes (or a single value) of price or cost, rate of employment, number of customers etc.

RIVET QUANT team conducts time series analysis based on (for example):

Linear Models
ETS models
ARIMA
ARCH
GARCH

05. Text Mining

Using text mining methods in data analysis has become the pillar of strategic brand management. Due to the increasing significance of brand loyalty, companies pay bigger attention to, for example, opinions about the brand appearing in social media.

Our specialists can provide text mining analysis using following methods:

bag of words
word2vector

06. Survival analysis

Survival analysis aggregates the set of statistical methods made for studying time which pass till the occurence of observed phenomenon. The methods allows us to conduct a forecast for different sectors. For example, we can forecast the policyholder's longevity or study the product durability.

Data Visualisation

Data Visualisation is very essential part of data analysis for Rivet Quant

We have noticed that Visualisation of data is the only thing needed by our clients to help them take important business decisions.
In the area of data Visualisation we can:

Design and implement dashboards

with effective and informative graphs.

Implement statistical analysis

in the dashboards.

Statistical products

The next core service which we offer our clients is optimisation of analytical processes.
In this area we can build e.g.:

Tools tailored to your needs
e.g. the tool which can be used to integrate different types of data.

Web application
based on e.g. Shiny packet

Knowledge Transfer

Rivet Quant members' long-standing experience in implementing analytical processes in different projects and sectors validates our comprehensive approach and opinion that the success of exploitation of analytical processes in organisations doesn't finish after building tools and presenting recommendations. We think that the success of analytical projects also depends on well-qualified staff. That is the reason why we offer you:

Trainings and workshops tailored to
your specific needs

Increasing staff competence
and skills

Support in hiring
qualified employees

Before organising the training we try to recognise participants' needs, define the level of expertise and desired scope of training as well as it is possible. We take such action to maximise effectiveness of training and possibility of usefulness of gained knowledge in participants' daily work.
Our training specialists are also very focused on the practical part and workshops while conducting the training.

Examples of training topics
• • •

Examples of training topics

* Broadly defined data analysis (measurement theory, probability theory, mathematical statistics, data analysis, data mining, effective data reporting and presentation, statistical package R, construction and validation of predictive models, CRISP-DM methodology CRISP-DM etc.).
* Basic methods for regression problems with generally available tools.

* The problem of prediction – examples of using available programming tools.

* Market basket analysis for the trade and e-commerce.

* Advanced problems of ‘Big Data' or ‘Deep Learning'.

* Methods of data clustering for their deep analysis and tools for detection.

About company

Rivet group is a team of specialists who change clients' needs into real solutions every day. We think out of the box and we connect different areas of business. Thanks to that, we don't limit our ideas to standard commercial solutions and our clients receive products characterised by the highest quality and tailored to their personalised needs. Our business model is best described by 'synergy'. We combine different competences depending on needs of client and project.

ABOUT RIVET GROUP

Rivet Quant is a part of Rivet Group
Copyright © Rivet Group
Designed by Black Flamingo black flamingo logo

This page contains icons from flaticon.com