At Twinshire we dive into many projects, from VR to Arduino
GUI Tool to look on your current local network for other computers to chat with. Using TCP sockets.
(Machine Learning - Labelled Data - Naive Bayes Score) This program is an introductory exerciser to Machine Learning. In short, the program scrapes IMDB for The Office Episode reviews, or any show/movie can be given. Then from this collected DataSet, its determined if it was a good or bad review by checking the rating the user gave (>= 8 stars is good), hence this is labelled data. The text/wording used in a review is organized into a category of good or bad based on classification of the review.
Now that certain words in both classes (good and bad review) have been collected from the labelled data, a Vocabulary (word frequency) then The classification Model (Compute the frequency and its conditional probability from the Vocabulary) can be build. Now the simple Machine Learning model has been trained with labelled data, new reviews are tested against the model to classify if they are good or bad reviews. Teir scores are calculated with the Naive Bayes Score.
Finally, the results of the model are outlined and detailed in the text files of the Repo. The quality of the model is measured by recall, precision, accuracy and f-measure.
This project allows for an Arduino robot (or any arduino board) to be controlled by a computer via Bluetooth.
This project contains Python code and Arduino code. The python code is responsible for detecting nearby bluetooth devices, then connecting to the selected arduino device via bluetooth and then sending data to it. The arduino code is reponsible for listening for incoming data and perform arduino actions in response, such as activating servos or motors to move.
But of course this project be instantly applied to communicate with any arduino board with a bluetooth module and not just an entire robot.