Neural Networks
Neural Networks代做 | security代做 | Network | Android | arm作业 | 代做network | Algorithm | bash代做 | app代写 | shell代做 | Machine learning代写 | 代做project | AI – 这是利用Neural Networks进行训练的代写, 对Neural Networks的流程进行训练解析, 包括了Neural Networks/security/Network/Android/arm/network/Algorithm/bash/app/shell/Machine learning/AI等方面, 这是值得参考的project代写的题目
- Using Beacons or RFID devices In this kind of project, you would use location finding using Beacons or RFID devices to track people in hazardous environments, nursing homes, hospitals, etc. These would be used in conjunction with embedded Linux boards acting as small servers at various locations. The Linux server will determine the coordinates of each tracking device and transmit the data (encrypted) to a central server (wifi or cell phone dongle), which will update a map display.
- Linux-based network implants These projects would use small Linux platforms (Pi, etc) as network implants. This is becoming a serious security issue in networks. The implants can be surreptitiously inserted into various servers/workstations and network switches and run malicious software to steal and exfiltrate information.
- Vulnerabilities in IoT devices and networks Given the proliferation of IoT in the use of home security and utilities infrastructure there is an increase in vulnerabilities and exploits in those networks. The main problem is a lack of proper firewalls and IDS/IPS in IoT devices.
- Software defined radio SDR (software defined radio) projects are currently very popular due to a drop in prices for powerful and small SDR units. These can be used in a very wide range of applications. The most popular is to detect intrusive drones in the vicinity and jam them within an operational area. The drones often interfere with public safety operations such as police incidents, fires, etc. For example, an SDR can capture wireless security window/door sensor signals and decode it using a (Pi/Arduino/FPGA). This technique replaces the traditional al arm central console.
- Custom penetration testing applications Designing and implementing custom penetration application for use with devices such as the Pineapple, shell 代做 script代写”> bash Bunny, etc. These are essentially small Linux machines but companies like Hak5 bundle pentesting applications with them. Developers can also design custom applications and install them on these devices.
- Computer Vision to Help Nature In this category of projects, you will use Yolo v5 (https://pjreddie.com/darknet/yolo/) to detect and classify various objects. The goal is to identify areas of nature that would benefit from AI and automatic detection/classification – for example, plastic detection in images in the ocean, whale detection to calculate the number of whales in specific seas, other animals, etc. Steps: i. find/create a dataset of images ii. annotate using Roboflow iii. train and test using the Yolo algorithm
iv. optimize/apply additional methods
v. write a report
- Search and Rescue The goal of this project is to create/improve algorithms related to drone-based Search and Rescue operations. You would explore topics related to path planning, computer vision, processing time optimization that can be later on utilized on DJI drones via SDK.
One of the possible ideas is to follow this competition and try to implement a solution
for it: https://www.computer.org/publications/tech-news/events/uav- 2022
Another is Path Planning: https://www.hindawi.com/journals/wcmc/2018/2851964/
The third one is object recognition: https://developer.dji.com/onboard-
sdk/documentation/sample-doc/advanced-sensing-object-detection.html
You can specify yours as well!
- Roster Scheduling (few projects, based on algorithms) The goal of this project is to apply algorithms to schedule employees with soft and hard constraints. You can select one of the following and then work with it:
a) Grey Wolf Optimization -> https://seyedalimirjalili.com/gwo |
https://github.com/7ossam81/EvoloPy
b) Whale Optimization -> https://seyedalimirjalili.com/woa |
https://github.com/7ossam81/EvoloPy/tree/master/optimizers
c) DragonFly Algorithm -> https://seyedalimirjalili.com/da
d) | https://github.com/wmalarski/DragonflyAlgorithmPy/blob/master/da/da.py
e) Moth Flame Optimization -> https://seyedalimirjalili.com/mfo
f) | https://github.com/7ossam81/EvoloPy/blob/master/optimizers/MFO.py
g) Sine Cosine -> https://seyedalimirjalili.com/sca |
https://github.com/7ossam81/EvoloPy/blob/master/optimizers/SCA.py
Steps:
i. identify type of employees (nurses/firefighters, etc.)
ii. learn about their shift restrictions
iii. apply chosen algorithm
iv. optimize/apply additional methods
v. write a report
- Networking (Edge Caching) In this project, you will use a simulator that was developed in the previous term to implement Machine learning 人工智能”> Neural Networks for offline/online (up to you) routing problems in Content
Delivery Network. The goal would be to optimize cloud traffic/perform traffic prediction,
etc. For example - Netflix delivers its content via CDNs. But the problem is how to
distribute the content properly between all the servers? Some of it needs to be cached -
but for how long? Your algorithms will try to find the best trade/offs, similar to the work
done here: https://www.researchgate.net/profile/Michal-
Aibin/publication/343994854_Content_Delivery_Networks_-_Q-
Learning_Approach_for_Optimization_of_the_Network_Cost_and_the_Cache_Hit_Rati
o/links/5f4d0d29299bf13c5067e247/Content-Delivery-Networks-Q-Learning-Approach-
for-Optimization-of-the-Network-Cost-and-the-Cache-Hit-Ratio.pdf
https://www.dragosilie.se/pubs/ilie2016-virtual_cdn-PREPRINT.pdf
https://www.mdpi.com/1424-8220/21/15/
- Machine learning in games These projects involve using implementing machine learning algorithms in games. These could include behaviour of computer-controlled enemies, dynamic terrain and level generation, or some other aspect of gameplay. You would have to generate training, testing and validation data, and show its effect in the game. If it does not improve the game, you would have to describe why. You would have to implement the algorithms and not just use off-the-shelf libraries blindly, and design and implement the game.
- Multimodal interface game Game interfaces can make a big difference in the quality of the game. In addition to joysticks, mouse and keyboard, many innovative games make use of other modalities. These could include haptic, eye or gaze tracking, head tracking, hand gestures, motion- capture or biometrics (e.g., using smartwatches). These projects would demonstrate some novel way of combining multiple input and output modalities. Alternatively, you could build a new game engine that can handle multiple input/output modalities and demonstrate it with a simple game.
- AR/VR in games AR and VR are gaining increased popularity in games. Building either an AR or VR game is not new, but there is plenty of room to create new and interesting games using the technology, especially in educational games (https://link.springer.com/article/10.1007/s11042- 020 – 09046 – 7 ). Alternatively, implementing new algorithms in AR (for example by combining it with deep learning https://ieeexplore.ieee.org/abstract/document/9447671) can lead to enhanced games.
- Realistic rendering in games While GPUs and graphics processing has come a long way, there are still certain things (e.g., hair, water, skin) that are difficult to render realistically. Implementing an algorithm that enhances realism and can be done in real-time (e.g., but leveraging high-
end GPUs or special modern processors such as the Apple M1 chip) can be explored and
demonstrated in a simple game as a proof-of-concept.
- Procedural terrain or level generation in games Procedural generation has been used for years in games, for example for terrain generation or level generation. However, complex terrain generation is still a difficult problem (https://ieeexplore.ieee.org/abstract/document/9232519), and there are possibilities to explore the use of deep learning and other machine learning methods in combination with procedural generation (https://link.springer.com/article/10.1007/s00521- 020 – 05383 – 8 ).
- Use of GPU in game engines As GPUs have become more and more powerful, including on mobile devices, their use both for advanced shaders and processing has become more possible (https://link.springer.com/chapter/10.1007/978- 3 – 030 – 59990 – 4_9). These projects would aim to either leverage GPUs directly to implement advanced shaders, or use them to implement certain parts of a game engine (e.g., physics or AI) more efficiently.
- Eye and gaze tracking for mobile applications Eye tracking (finding eyes in a video stream) or gaze tracking (estimating where a person is looking through tracking eyes and other features in a video stream) have numerous applications. Their effective implementation on mobile devices is yet to be achieved well. Projects in this area could explore interesting applications of either type of tracking using cameras on a mobile device. One of these could be adding eye tracking to an existing app developed for human activity monitoring using Android devices. Other projects could explore the use of existing eye or gaze tracking devices to enhance existing apps, including games.
- Ubiquitous mobile human activity monitoring A context-aware Android data collection app has been developed at BCIT, with the aim of driving empirically-based wheelchair design and use analysis. Currently it is in need of updating and enhancements, including revamping the user interface and adding other sensory input. In addition, it can be extended and used in the context of monitoring of urban travel behaviors, mental health and smartphone use among adolescents.
- Educational device prototyping With the recent coronavirus pandemic, the use of digital devices for education became critical. Devices designed primarily for entertainment and other uses (many of which are designed to distract and addict users) began to be used for education, with mixed results. These projects would aim to modify existing devices and apps, or prototype new devices and apps, that are designed fundamentally for educational activities.