Throughout my studies my research has covered some pretty big topics giving me the opportunity to deep dive into many areas I feel are highly applicable to the future of communication. This research culminated in an offer for publishing my Capstone work.
I achieved a mark of 95% (High Distinction). I am very proud of this achievement. This has led to publication of my work which is in process. As such my research paper is not made available for viewing on this platform.
Built a Gradient Tree Boosting model machine learning model for predicting the average price (or direction of pricing trends) of wine produced from the McLaren Vale wine region.
Build a layered BlockChain for data security as a digital twin of a specific vineyard, the cases of wine produced for sale by the vineyard and the bottles produced within the cases of the wine.
Built a system of verification, authentication and security for the wine producer and consumer through the implementation of Quick Response (QR) codes.
What Block Chain is and how a distributed ledger is imperative to logistics and security for the future of wine. I research how blockchain could be applied to the wine industry. Particular interest was paid to the creation of Digital Twins for wine regions, vineyards, and vintages.
I built a machine learning model derived from collected & publicly available data pertaining to weather, production, wine rating and existing historical wine pricing. My algorithm uses differing test cases to build the model picking the most accurate each time to incorporate into the machine learning model.
Gradient Tree Boosting
I focused on gradient tree boosting algorithms to build my machine learning model. This was a focus on weak learners to build an accurate prediction. My solution did compare and contrast differing algorithms for linear regression along with various Gradient Tree Boosting Algorithms to find the optimum algorithm.
As I have experience in the wine industry as a Sommelier in Sydney I had a great opportunity to engage with winemakers, agronomists, viticulturalists and horticultural academics to find out their concerns. I realised there are a large number of stakeholders at differing levels of the winemaking process invested in the success searching for the holy grail of grape yield prediction.
COVID-19 thwarted my plans to help with vintage while researching processes and interviewing stakeholders. Engagement was undertaken online although I would have loved to introduce my daughter to the land and people that care for it.
Embedded software was researched extensively with the idea that data capture is of paramount importance to my projects. Getting and maintaining good data from embedded devices over a period of time will allow for better more accurate information and machine learning.
The Internet of Things (IoT) plays a huge role in the collection of data. Research into how embedded devices in the IoT can be employed in the wine industry, what already exists within the IoT solution scope for the wine industry and what possible solutions could be employed to help the wine industry not only in the growing of grapes but the sale and distribution of wine products.
Integrating the tiered blockchain system I designed and implemented into a system for verification and authentication of saleable products through Quick Response (QR) code generation and implementation. This security of distributed data and delivery not only protects the investment of stakeholders for consumption and sale but top level investment of the vineyard as an operating business. This is done through tracking of weather patterns and production processed through my algorithm.