Você procurou por: how much string is needed to go around the bucket (Inglês - Persa)

is it fun to go around the world as you thought in the palace?
همون طور که در قصر فکرمیکردی گشتن دور دنیا با مزه هست؟

and to give you a sense of how short that is, the sun takes 200 million years to go around the center of the galaxy.
و برای اینکه به شما درکی از میزان اندکی این مقدار بدهم برای خورشید ما 200 میلیون سال طول می کشد تا به دور مرکز کهکشانمان بچرخد

i don't think it's good to go around the problem with silla princess.
فكر نمي كنم ملاقات پي در پي با پرنسس شيلا كار مناسبي باشه

and then the last section was going to go around the rail yards, which is the largest undeveloped site in manhattan.
و آخرین قطعه در اطراف منطقه راه آهن خواهد بود، که توسعه نیافته ترین منطقه منهتن است.

i began talking with people who actually had been out to the gyre and were studying the plastic problem in the marine environment and upon doing so, i realized actually that cleaning it up would be a very small drop in the bucket relative to how much is being generated every day around the world, and that actually i needed to back up and look at the bigger picture.
با افرادی که در باره‌ی آلوده‌گی پلاستیکی در محیط دریایی مطالعه داشتند شروع به صحبت کردم وبعد از اون صحبت ها متوجه شدم با توجه به میزان آلود‌گی که هر روز در جهان داره تولید می‌شه تمیز کردنِ این آلود‌گی اثرِ چندانی نداره. متوجه شدم که باید نگاه کلان‌تری به موضوع داشته‌ باشم.

if we could get exquisite control over the essence of energy, so i started to go around the world finding the best and brightest scientists i could at universities whose collective discoveries have the chance to take us there, and we formed a company to build on their extraordinary ideas.
و چه میشد اگر مامیتوانستیم کنترل دقیق گوهرۀ انرژی ،الکترون ، را بدست بگیریم. و چه میشد اگر مامیتوانستیم کنترل دقیق گوهرۀ انرژی ،الکترون ، را بدست بگیریم. خُب من شروع کردم به سراسر جهان رفتن ، تا بهترین و درخشانترین دانشمندان را خُب من شروع کردم به سراسر جهان رفتن ، تا بهترین و درخشانترین دانشمندان را خُب من شروع کردم به سراسر جهان رفتن ، تا بهترین و درخشانترین دانشمندان را در دانشگاها بیابم که اکتشافات جمع آوری شده آنها، شانسی برای بردن ما به آنجا بود، در دانشگاها بیابم که اکتشافات جمع آوری شده آنها، شانسی برای بردن ما به آنجا بود، ما یک شرکت برای ساخت این ایده های فوق العاده ایجاد کردیم. ما یک شرکت برای ساخت این ایده های فوق العاده ایجاد کردیم.

the case study followed a methodology that included an exploratory phase conducted to select and engineer potential features, the development of a performance metric in alignment with the specific goals of the case study and the creation of an experimental design to systematically analyse the success rate of different algorithms. our results indicate that adding engineered features in the data, namely agility, outperformed other experiments leading to the final solution. an important contribution is the novel application of machine learning in predicting supply disruptions. our results are promising as they present a significant improvement in the prediction of disruptions with limited internal historical data available to the oem. however, a key learning is that the application of machine learning in this particular field presents several challenges. firstoftheseisthatadisruptionproblemintheindustrybydefinitionpresentsclassimbalance:therearefewerdisrupted orders than successful orders. similar issues have been observed in other industrial machine learning problems such as quality issues on a production line. secondly, disruptions on supply might be dependent on a variety of external factors such as traffic, weather, machine breakdowns and thus their combination might present seemingly random patterns when using only internally available data. complimentary data from external, publicly available sources or supply chain partners might help increase prediction performance. the third issue is the curse of dimensionality. when the number of variables in the feature space is high, data relating to each attribute becomes sparse, hindering statistical significance. we tried to mitigate this issue by using a cut off value of five samples of each attribute during the pre-processing stage, however, more experimentation may be necessary. we have shown that it is feasible to generate augmented features that improve the results of the predictions. it appears that agility in this case had a positive impact on performance. this feature might not be appropriate for other industrial settingsormightneedtobefurtherfine-tunedusingproductionvolumedata.similarly,itsoptimaltimewindowmayalsobe different from case to case. teasing out appropriate features from data and optimising their use requires domain knowledge, which we argue should underpin the machine learning process in industrial applications such as the one presented in this paper. ourcasestudydomain,acomplexassetmanufacturingcompany,mayinduceotherlimitationstogeneralisationtowards different industries, as supplier delays and their management may be differently assessed. as with single case studies in general, the selected company may possess inherent characteristics strongly different to other companies. in addition, the timeframe of the analysis may describe a specific state of the supply chain. several further avenues of research needs to be taken. the focus of dataset augmentation has relied on constructed features, excluding the use of external data. this provides an opportunity for future research. for example, some preliminary work has been done around the extraction of the localisation of suppliers from public sources of data, and the creation of a feature relying on this information, potentially allowing the learning algorithm to deduce location-based relationships. additionally, further research is needed to improve the performance of predictions, by testing new methods and optimising the learning process. for example, the confidence level of predictions could be used to generate predictions for orders within certain confidence only, further improving the precision and recall. the classification-based approach we have taken has relied on well-known classification algorithms and a frequentist approach. other approaches such as bayesian models, or decision trees may be worth experimenting with as they might be useful in estimating underlying probability distributions of the feature space.
چندین راه تحقیق دیگر باید مورد استفاده قرار گیرد. تمرکز افزایش داده ها به استثنای استفاده از داده های خارجی به ویژگی های ساخته شده متکی است. این فرصتی را برای تحقیقات آینده فراهم می کند. به عنوان مثال ، برخی کارهای مقدماتی در مورد استخراج محلی سازی تأمین کنندگان از منابع عمومی داده و ایجاد ویژگی متکی به این اطلاعات انجام شده است ، که به طور بالقوه به الگوریتم یادگیری اجازه می دهد روابط مبتنی بر مکان را استنباط کند.