added in photos, added in CS side of deliverable 4, edited my blog 3 …

…to add in contributions

content/FraniPendus/blog3.md

@@ -27,3 +27,11 @@ After the hike, we crossed the bridge from Schengen, Luxembourg to Perl, Germany
 Sunday, I went to Lake Echternach with the intention to swim in it. However, upon arriving, I realized I really did not want to swim in that water as I could see a plethora of bugs and algae. Regardless, I got to walk around the lake and took in the views of nature once again.
 
 Monday, we finally returned back to Leuven. Even though I was so sad to say bye to Luxembourg, I know it is a place I will definitely go back to.
+
+In terms of the project, I created all of the tables and generated all of the fake data. I added them into the database section on VS Code so that I could begin write the api routes. I wrote 12 api routes:
+* 1 x 'put'
+* 1 x 'post'
+* 1 x 'delete'
+* 9 x 'get'
+In my opinion, the 'get' and 'delete' routes were easy to implement; however, the 'post' and 'put' were VERY difficult for me and required a lot of time due to the string formatting, specifically pertaining to dates. It took me around 3 hours of staring at the code and getting help to realize I was missing a couple *single* quotation marks. 
+In the end, it did work out and it was so satisfying to see the code I wrote "behind the scenes" performing tasks the user wanted. 

content/team_posts/deliverable4.md

@@ -0,0 +1,39 @@
+---
+title: "Project - Deliverable 4"
+date: 2024-05-27
+draft: false
+description: "Final Project!"
+tags: ["authors", "config", "docs"]
+authors:
+  - "nalika_palayoor"
+  - "erica_indman"
+  - "frani_pendus"
+showAuthorsBadges : false
+---
+
+# Project update 4 
+
+## **Machine Learning Models:**
+
+## **Software Architecture:**
+![](arch.png)
+
+Our database, RoundTrip, has 19 tables and more than 3000 data entries. Our middleware, the REST API built with Flask, has 4 main sections: machine learning model, travelers, advertisers, and deal administrators. The machine learning model contains the ML model routes, as well as the functions for each linear regression prediction (including the functions to train the models). The ML model routes are used in the front end by both the System Admin who trains the models and the travelers who use the models to predict the flight and hotel costs. The travelers middleware section contains the routes that execute the actions the travelers want to perform (aside from cost predictions). The advertisers middleware section contains the routes that execute the actions the advertisers want to perform. The deal administrators middleware section contains the routes that execute the actions the deal administrators want to perform.  
+
+## **Database Model:**
+*The initial database model* : 
+![](global.png)
+
+*The final database model*: 
+![](finaldbmodel.png)
+
+As I began to implement the app’s functionality, the model had to change. The exchange rates attribute became its own table but was later deleted since it was no longer used for machine learning predictions. The favorite flights (’favFlights’) table was also deleted for lack of impact on the UI experience since, as a frequent traveler myself, I’ve never had a favorite flight. Finally, 3 new tables were added: 
+    - the main data frame (‘main_df’)
+    - hotel parameters (‘hotel_params’)
+    - flight parameters (‘flight_params’)
+These tables are necessary for the machine learning models. The main data frame was needed for model training as it allows us to get the data needed to perform the linear regression and come up with the line of best fit for predictions. The parameters tables store the coefficients from both lines of best fit (1 table per prediction type, ie. flight prediction has its linear regression parameters stored in ‘flight_params’) and during the prediction allow us to retrieve these coefficients to apply to a new set of inputs provided by the user. 
+
+
+### Presentation Link:
+https://docs.google.com/presentation/d/19QlnvmbjLDo0bogDgHAWGeKj8quzH5o8jgYfyQWjT7g/edit?usp=sharing
+