Improved Decisions for Ocean FreightsProject Management MethodologyPlan
Data Collection |
Data Cleaning and Initial Visualisations |
1 Iteration per question:
Building of Visualisations |
Dashboard Design |
Data Cleaning and Initial Visualizations
We will need to study the data and identify which columns of data is not useful for analysis. We will also need to filter out noise. We will then come up with some visualizations using JMP first to explore which visualizations have the potential to be useful. We will not be using JMP for our final dashboard as JMP can only produce static visualizations whereas our final product will contain interactive visualizations. Calculated fields will also be explored. Depending on the data, we will decide what calculated fields would be useful for use in our dashboard.
Further Data-Analysis
We will attempt to see if the data can be clustered. If meaningful clusters can be formed, we can use this information to help in our dashboard design as well. We will also check for relationships between variables by looking at measures such as correlation.
Building of Visualizations
Relevance, trends, and concepts are hard to understand just from spreadsheets of data, and as such, visualisations are critical in conveying these messages. It is our aim to not just present the findings in charts and graphs but also in interactive visualisations. Aligned with our client’s requirements, we will be looking into creating a sustainable solution, one that can be used repeatedly.
Dashboard Design
We will carry out an iterative approach on different dashboard designs to see which design gives the most insights. We will then attempt to incorporate what-if analysis into our dashboard. We will be studying these areas:
- What type of visualizations to use
- What selections and filters are most effective
- Level of customizability
- Colour schemes and layout for clarity