面试经验可以在一定程度上弥补我们先天在语言与文化差异上的不足，并且能给招聘方一个优秀的第一印象，从而获得更多的机会。另外，良好的投简历策略也可以帮助我们更有效率的与招聘方沟通，同时节省宝贵的时间。
本文着重讨论投简历策略，基础及职位面经网上到处都是，写得都蛮好，这里就不重复了。

知乎上有一个贴叫“英国找工作到底有多难”，里面提供了大量的网站资源与经验之谈，对我的求职之路也有很大的帮助。这篇主要细说一下实用网站的用法，以及个人对recuiter和Guaduate Scheme的一些看法，希望能帮到大家。

老规矩，未经授权严禁转载。

I feel Spanish sounds like Japanese guy speak English xD.

转眼已经在英国工作了半年，身边开始找工作的朋友后辈也逐渐变多，于是有了把各种经验写下来的想法，一来可以直接发给朋友，省事。二来也许可以帮到更多的人。

Datetime format is very frequently used in real world. This post summarized basic usage of datetime in python and pandas and introduced the dateutil library with examples.

As a Non-EEA citizen without the work permit, I faced the most desperate situation during the job searching in my life (without one of): nearly 150 applies, and more than 90% of them never reply because I don’t have the work visa. So, what is the work visa? And how difficult is it for a non-EEA foreigner to get a job in UK without the work visa?

Hope this (and following) post can help the people who are also struggling and dreaming to work in UK because of any motivation, interest and fever.

The first and foremost advice: inform your employer at least 3 months in advance to process your sponsorship number.

Echart is an amazing D3-based visualization plug-in for the Hexo. Here is a sample rendered by Echart.

These days I’m playing a simple but interesting Pokemon with stats dataset. Unexpectedlly,the seaborn and matplotlib do not support the radar chart. And the matplotlib radar chart sample is totally a mass (186 lines, what the hell). After an afternoon’s work I implemented the radar chart in a much simple way (only 20 lines). Hope this blog could help someone who is try to draw the radar chart in Python.
(Or time to change to R? ;P)

CUHK-03 dataset is a baseline dataset for person re-identification tasks, published on 2014. The original file is a 1.73GB .mat file which compress all images and testsets settings together. Because my project is mainly based on Python and Tensorflow/Keras. Therefore, I need to decompose the original file to .jpg, then retrain them in the GoogLeNet by Tensorflow.

I searched on the Internet before and there is no specific article about how to extract images from the mat by Python. Hope this blog can help the people who are interested in person re-id as well ;).

Titanic is similar to the problem in practical. Each feature has it own meaning and background, which can help us to build a better model. I also acquired knowledge about the history of Titanic and traditional prefix culture in English.

If I have time, I have to been to the Merseyside Maritime Museum again and I’m sure I can gain some new and unique experience from there.

Current Score: 0.81340 (Top 7%)