Reference no: EM133123046

Use Case: Third-Party Support

The application created through this project should allow other developers to utilize an API to build their own applications. The idea is that other developers could then use the inventory management system as a service to enable their application. For example, this system could expand to inventorying food products in a fridge. An external application could then use the API to pull data about the food products that the user has available to produce a grocery list or a recipe. Another option is identifying items through an image and having a robot perform something else.
The third-party developer would expect the following:
• Access to inventory details through an API server.
• The ability to send an image to an application server and receive a response containing machine learning predictions of items within the image, such as:
? Text detection
? QR/Barcode detection
? Object position
• Application access to authenticated APIs
These two use cases are the foundation for the application's success. Without inventorying RAM sticks, the proof of concept would not be strong enough to ensure a viable product, and developing the application for third-party support shows scalability.
Top-Down Approach
The top-down design approach begins with a listing of the primary functions of the system. For example, if the software has three primary functions, then each function is treated as a major task. The major tasks are then analyzed and further decomposed into subtasks as needed.
Use the top-down approach in the analysis of the system requirements. This will then be used in the design of the system. Represent the top-down design with a diagram similar to a flow chart. A typical top-down approach in data science is:
1. Formulate a hypothesis about the problem
2. Collect data
3. Build a model
4. Deploy the model
5. Test the hypothesis
6. Inform decision makers about the characteristics of an action plan

Technical Requirements
Provide a listing of the technical requirements that the system must fulfill. List these requirements and include brief descriptions. Follow the listing of technical requirements with a listing and explanation of the software and/or hardware that is necessary to meet the requirements.
Data Modeling and Analysis Process
Provide diagrams illustrating the type of database or data types and format. The data should reflect the requirements of the problem/challenge being addressed. Include naming conventions and data integrity; data archiving/retention; privacy and security information; data replication; data flow diagrams (DFD) and pipelines; statistical models; analytics models; data mining models; machine learning models.

The main steps in building a data science model are:
• Define objectives
• Collaborate with stakeholders
• Collect data for exploratory analysis
• Determine the functional form of the model
• Split the data into training and validation
• Assess the model performance
• Deploy the model
• Improve the model

Provide a draft of each user interface screen required by the system. These interface "sketches" will be refined in Milestone 2. Diagrams should include components such as, but are not limited to:
1. Title
2. Toolbars
3. Date entry fields
4. Content, such as text, charts, tables, etc.