Adjust examples to new `@initialization` macro

[wouterwln]

Hello everyone, thanks for revamping the examples for the new release of RxInfer. Unfortunately we have another breaking change that requires a change in the examples. Luckily, this only affects the examples where we need to initialize marginals or messages. Since we're aiming to release on very short notice, we'll be employing @albertpod's favorite strategy: hostile assignment. I've made an overview of which examples need change, and I'll write a migration guide for the new syntax (ready in #278 ). This should't take more than 5 minutes per example, since most code can also be found in the tests. The assignment is merely for everyone to see the new syntax in action once before we release. I'll link the migration guide as soon as I have it ready, but if you check the tests, the code should be very straightforward.

• [x] Initialization page in the manual @wouterwln #278

Basic Examples

• [x] Bayesian Linear Regression @HoangMHNguyen
• [x] Kalman Filtering and Smoothing @Sepideh-Adamiat
• [x] Predicting Bike Rental demand @abpolym
• [x] HIdden Markov Model @mhidalgoaraya

Advanced Examples

• [x] Advanced tutorial @wouterwln #280
• [x] CVI @Nimrais
• [x] Infinite data stream @Raphael-Tresor

The rest I'll fix myself #280

[wouterwln]

Update: I'll fix all of them, thanks @HoangMHNguyen !