Before seeing a lot of friends who want to change careers, they will ask, for example, whether girls are suitable for data analysis, whether liberal arts students can do data analysis well, what data analysis can do, and so on.



What I want to write today is a summary of the past three years, and it is also for those students who want to engage in data analysis. I hope it will be helpful to you from a rookie to an entry-level three years.

说到开始接触Python应该是2016年12月26日(刚刚考完研),这一天主要做的就是安装Python,当时安装的是Python的原生包,麻烦一些,在配置环境变量的时候一直出错,结果安装了两天,才打印出了Hello World!,其实说真的自己差点放弃。



Later, I slowly felt the fun, one of which is very important to figure out the problem first, which requires us to decompose the problem step by step, and then use search tools to solve it step by step. In fact, in this process, the train of thought will become clearer and clearer, and finally more than half of the problem can be solved, and it can be very efficient to ask others at this time.



So when you are learning, you must often use the search tool (Google). Every time you search, you will be able to improve your ability to search keywords. On the other hand, when you ask others for advice, others can easily ask and answer your questions and think of others. In fact, you are thinking about yourself, isn't it?






如果大家打算写一个爬虫程序作为自己找工作的项目实践,推荐使用requests库和Beautiful Soup,再加上QQ浏览器上扩展程序XPath Helper,将会是非常的方便。有一点还请注意,这是一个LINUX下的教程,关于学习爬虫方面,其实涉及到的LINUX命令只有进入一个文件进行编写,复制一个文件,简单的几个命令。




When I first started learning, my goal was to copy down other people's code, as long as I didn't report an error. At that time, I had a very bad habit of copying the code and never writing the code by hand. Always want to learn quickly, in fact, without the precipitation of their own thinking, in less than two days will forget the logic of the code. Therefore, it is also recommended that you can write it yourself, and then compare it with other people's code to modify it.



Learning programming, really do not be afraid of trouble, there will be mistakes, but as long as you can find out the causes of errors, a little bit of accumulation, it must be a great progress.





Generally speaking, large companies are very demanding, and they can learn a lot from the technical direction, and the treatment is also very good. If a small company claims something like big data, artificial intelligence and so on, ha ha, it is either deceiving users or investors.



Believe me, as long as you choose the Internet, whether you do technology or product or operation or data analysis, you can't escape middle-aged anxiety. If you are afraid of this, choose the traditional industry. But as long as you do your own thing well, there must be no other way, and it is not a great crime to be over thirty and not to die.



For beginner children, do a good job of their own work, leaders must be taken seriously, and timely feedback is very important. Think of ways to improve efficiency in your spare time, such as the form you just made, whether there is a faster way to achieve it, and whether the picture can be done more clearly and beautifully. At this time, you will not assume the role of data analysis and guidance of the department, and the boss will not attach great importance to your opinions, but slowly shine, and what you do will brighten the eyes of others, and presumably the leader will notice you.



Do not be stingy to help others, often help colleagues, you may have new knowledge to learn. And you helped him, and the next time you have a problem, I'm sure he won't flatly refuse you.




For the advanced, I am also groping. At present, there is no problem with the processing and expression of the data, which is enough to cope with the existing work. But how to extract information from the data to improve the business and optimize the process. I am still learning this point. From the performance of the existing data, it is very challenging to put forward constructive suggestions to operators, products and leaders, not only to ensure that the data are accurate (so that we can draw an unbiased conclusion). You also have to be familiar with business products, as well as an overall view of the industry.


文章来源:网络 作者: 网络 版权归原作者所有



The above content is not used for commercial purposes. If intellectual property issues are involved, please contact the editor and we will deal with them immediately.

Original: https://blog.csdn.net/xuezhangmen/article/details/123865171
Author: 学掌门
Title: 做了三年数据分析,给你的几点建议

PPASR中文语音识别(入门级) 人工智能


PPASR语音识别(入门级) 本项目将分三个阶段分支,分别是入门级、进阶级和应用级分支,当前为入门级,随着级别的提升,识别准确率也随之提升,也更适合实际项目使用,敬请关注! PPASR基于Paddle...
CTC算法详解 人工智能


目录 1.语音识别的基本架构 2. RNN循环神经网络 3.损失函数 4. CTC Loss函数计算 5. CTC训练算法总结 6. CTC解码 6.1 CTC Prefix Search Decod...