Announcing The Winners of Machine Learning for Transportation Risk Mitigation Data Challenge
Purpose of Challenge
The objective of this data challenge was to develop machine learning models that accurately assess risk in supply chain transportation vehicles of perishable goods such as livestock and food items.
It’s summertime and humidity is a problem in most of the western world this summer. This data challenge sought to understand the impact temperature and humidity have on the lifecycle of a given supply chain.
Supply chain management has been in flames in recent years. This challenge scratched the surface to learn how machine learning and artificial intelligence streamline and improve efficiency and sustainable practice in the transportation of perishable goods and animals.
Winners are as follows:
First place: Mohammad Jamali ($2500 Prize)
Mohammed took a unique approach to analyze the dataset provided and drew conclusions that were not initially thought of at the origin of this experiment hosted by Ocean Protocol. Mohammad's report included in-depth heat mats that helped us understand the geolocation and internal weather conditions in a new light. This paper sought the best way to find the correlation between the timestamp and temperature to approach the problem as a classification task. This was done by segmenting the timestamp as an example into 24 one-hour intervals or classes and assigning temperature values to these segments. By doing so, it became possible to estimate the likelihood of specific temperature ranges occurring within each time segment.
Second place: Marco Rodrigues ($1500 Prize)
Marco dug deep through the data set to prove how some hours of the day are warmer than others. This makes perfect sense and is verified by sound data regression models. This metric became interesting when comparing external weather temperature to the temperature inside the shipping vehicle temperature used for this data challenge. Marco’s report used sound data polishing tactics, extensive diagrams, and data visualization. He also used the data set to expose natural weather humidity levels in the context of this challenge.
Third Place: Andrey Bessalov ($1000 Prize)
Andrey built cutting-edge clusters by using weather features and concluded the best factors concluding the ideal conditions inside the vehicle. They are medium temperature, medium humidity, low wind, no snow, no rain. Andrey cleaned and visualized the dataset provided in terrific detail. From this report, we can assume the best conditions and humidity levels for shipping perishable goods.
There were 12 total submissions to this challenge, less than previous challenges which reach between 20–40 submissions. This challenge was a bit more challenging than others with the use of in-vehicle sensor data, IoT data, temperature and humidity metrics. Specs on the challenge itself can be found here.
Have you ever wondered how AI and ML can impact a real-world business case like supply chain management? This challenge was an indicator of that. Or, curious about how the performance of your supply chain is mitigating risk in handling shipping logistics and humidity? Let us know in our community channels, and your protocol may be selected next to run a data challenge on.
Congratulations to the winners, and thank you to all the participants for their hard work and dedication. This challenge wouldn’t have been possible without you. Stay tuned for our next stimulating data challenge!
About Ocean Protocol
Ocean was founded to level the playing field for AI and data. Ocean tools enable people to privately & securely publish, exchange, and consume data.
Follow Ocean on Twitter or Telegram to keep up to date. Chat directly with the Ocean community on Discord. Or, track Ocean progress directly on GitHub.
