We Will Propose The Optimum Solution According To Your Data EnvironmentWe Will Propose The Optimum Solution According To Your Data Environment
- Get some extent of data
- Data is stored in the database
- Basic statistics by SQL etc. are made
- Some insights can be gained from statistics
- Basic statistics are substantial and can be managed by BI tools
- Manual exploratory data analysis can be performed using a pivot table, etc.
- Measures can be planned from manual data mining
- Exploratory data analysis using machine learning algorithms
- Make data-based decisions using A/ B testing, etc.
Providing infrastructure
- There is an environment where measures can be automatically executed using machine learning, etc.
- There is a mechanism to make stable profits by machine learning
- Build more advanced machine learning algorithms (such as deep learning) to further improve profitability
Features Of The ServiceFeatures Of The Service
Full support from analysis to solution construction and operation
Experienced and proven
expert team
Advanced technology know-how and abundant staff
Analyze Accumulated Data To Derive Valuable Output
You have data available, but don’t know how to use it. You don’t have staff who can analyze the data. You want to carry out marketing activities based on data. VIETIS will help you solve “ Increasing sales” , “Expanding profit” and “Making speedy decisions”.
Issue
Investigate issues
- Business analysis
- Identifying and classifying issues
- Prioritization
- Data collection and preprocessing
- Data cleansing
We perform market research and analyze business/ business flows to identify issues. After further classifying and prioritizing the extracted issues, we will collect sample data.
Data analysis
Design data model
- Model building
- Model performance evaluation
- Parameter adjustment
- Comparison and selection by multiple models
This is the process of creating a data model in order to clarify data requirements and determine the implementation range. Start with simple machine learning, verify performance, and decide which model to use.
Solve business
issues
Build system
- Data collection platform building
- Operation management tool building
- Machine learning tuning
- Production introduction.
We will introduce a system that automates the process of data collection and data preprocessing, as well as management tools for verifying and evaluating the results. At the same time, we will change the algorithm and build an environment that optimizes machine learning to improve the accuracy of data analysis.
Case StudyCase Study
Analyze about 20,000 DATA flights and establish aviation fuel-saving solution
Problem
The client spent about 77 billion yen on fuel costs in 2017. This accounts for 26.5% of total spending. In this project, we aimed to devise and implement a fuel-saving solution by data analysis.
Input
Fuel costs are affected by a variety of factors, including aircraft design, flight mode, load weight, and fuel volume.
The project team analyzed data from about 20,000 A350 aircraft for the two years 2017-2019, provided by many systems such as CFP, ACAR, and LOOMS.
Output
As a result of data analysis, it was discovered that there was a discrepancy between the amount of fuel required by the captain before the flight and the actual amount of consumption. As a result of changing the guidelines to require declaration when the fuel of 500 liters or more is required, almost no captain demands a fuel amount of 500 liters or more, and we succeeded in saving a lot of fuel.
