Group 10

In this section we will be discussing the databases we have created that will be utilized for our machine learning model. To get an idea as to what our database will look like once joined, the following is a snapshot of our ERD model and Joined table once the query is executed. (Click images below for larger view)

                    Select *
                    from seasons_and_quarters as sq
                    left join aapl_stock on
                    aapl_stock.date = sq.date 
                    order by sq.date;
                

Entity Relationship Diagram

These photos represent our most successful Machine Learning Model - the Random Forest Model. In this model, all price variable change columns (24hr changes in the low, high, open, and close prices) were incorporated, and the volumn’s 24hr change column was used to create its respective gain_loss column. Both the volumn’s change column and gain_loss column were dropped, and the gain_loss column was applied as the model’s target - resulting in the 72% accuracy.

The following image represents how each feature contributed to the model’s accuracy by weight of it’s importance.

The R-Analysis is used to show the apple stock volume changes and how it was affected by seasons and quarters. To better understand the analysis, we clearly used colors and titles to depicts what the graphs describes.

---

The diagram below shows the summary of the price volume of apple stocks for 40 years in relative to seasons. According to graph, the fall season has the highest volume accross the years and the volume increased drastically in the 20s

---

The diagram below shows the summary of the price volume of apple stocks for 40 years in relative to quarters. According to graph, the 1st Quarter has the highest volume accross the years and the volume increased drastically in the 20s

---

To give a more detailed explanation, we decided to selected the last five years(2016-2020) to predict the stock behaivor in the next five years. According to diagram, it shows that the winter period has the more volume in apple stocks until 2020. In 2020, we understood that due to covid; people did not trade during the winter period so the spring period took the lead here. It is safe to say that the winter period is the ideal season to invest in apple stocks

---

Tableau Analysis: For indepth analysis, we bulit a dashboard using Tableau. (Click on the image below to view the Dashboard)

---

See more here

• Data source: The data utilized is from Kaggle.com, titled the “NYSE”
• Bootcamp Modules: https://courses.bootcampspot.com/
• Stack Overflow: https://stackoverflow.com/

Volume was significant to accuracy of Gain/Loss predictions. When combined with volume, price variables had a higher accuracy than without volume consideration, and the “Open” price variable weighed the most in importance of all price variables.

The conclusion of our analysis depicts that opportunity for gains were most significant in the summer and during the 3rd Fiscal Quarter and 2nd Fiscal Quarter.

Moments of the greatest losses occurred in the winter season nad tie between the 1st and 4th fiscal Quarter.